From 4b9fd8cf6626727d5646d11642954880a7b3d096 Mon Sep 17 00:00:00 2001
From: johannes wasmer <johannes.wasmer@gmail.com>
Date: Mon, 13 Mar 2023 04:00:11 +0100
Subject: [PATCH] update bibliography
---
bib/bibliography.bib | 2347 ++++++++++++++++++++++++++++--------------
1 file changed, 1562 insertions(+), 785 deletions(-)
diff --git a/bib/bibliography.bib b/bib/bibliography.bib
index 182782e..518a6ad 100644
--- a/bib/bibliography.bib
+++ b/bib/bibliography.bib
@@ -26,7 +26,7 @@
urldate = {2021-05-15},
abstract = {Materials design and discovery are often hampered by the slow pace and materials and human costs associated with Edisonian trial-and-error screening approaches. Recent advances in computational power, theoretical methods, and data science techniques, however, are being manifest in a convergence of these tools to enable in silico materials discovery. Here, we present the development and deployment of computational materials data and data analytic approaches for crystalline organic semiconductors. The OCELOT (Organic Crystals in Electronic and Light-Oriented Technologies) infrastructure, consisting of a Python-based OCELOT application programming interface and OCELOT database, is designed to enable rapid materials exploration. The database contains a descriptor-based schema for high-throughput calculations that have been implemented on more than 56 000 experimental crystal structures derived from 47 000 distinct molecular structures. OCELOT is open-access and accessible via a web-user interface at https://oscar.as.uky.edu.},
keywords = {chemistry,descriptors,materials database,molecules,organic chemistry,visualization},
- file = {/home/johannes/Nextcloud/Zotero/Ai et al_2021_OCELOT.pdf;/home/johannes/Zotero/storage/DW64W25V/5.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Ai et al_2021_OCELOT.pdf;/Users/wasmer/Zotero/storage/DW64W25V/5.html}
}
@article{alberi2019MaterialsDesign2018,
@@ -46,7 +46,7 @@
abstract = {Advances in renewable and sustainable energy technologies critically depend on our ability to design and realize materials with optimal properties. Materials discovery and design efforts ideally involve close coupling between materials prediction, synthesis and characterization. The increased use of computational tools, the generation of materials databases, and advances in experimental methods have substantially accelerated these activities. It is therefore an opportune time to consider future prospects for materials by design approaches. The purpose of this Roadmap is to present an overview of the current state of computational materials prediction, synthesis and characterization approaches, materials design needs for various technologies, and future challenges and opportunities that must be addressed. The various perspectives cover topics on computational techniques, validation, materials databases, materials informatics, high-throughput combinatorial methods, advanced characterization approaches, and materials design issues in thermoelectrics, photovoltaics, solid state lighting, catalysts, batteries, metal alloys, complex oxides and transparent conducting materials. It is our hope that this Roadmap will guide researchers and funding agencies in identifying new prospects for materials design.},
langid = {english},
keywords = {DFT,materials database,MD,multiscale,roadmap},
- file = {/home/johannes/Nextcloud/Zotero/Alberi et al_2018_The 2019 materials by design roadmap.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Alberi et al_2018_The 2019 materials by design roadmap.pdf}
}
@article{amorosoInterplaySingleIonTwoIon2021,
@@ -65,7 +65,7 @@
issue = {8},
langid = {english},
keywords = {atomic scale magnetic properties,magnetic interactions,magnetic nanostructures,pgi-1 seminar,skyrmions,topological spin textures,topology and magnetism},
- file = {/home/johannes/Nextcloud/Zotero/Amoroso et al_2021_Interplay between Single-Ion and Two-Ion Anisotropies in Frustrated 2D.pdf;/home/johannes/Zotero/storage/FN7Y4K4H/htm.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Amoroso et al_2021_Interplay between Single-Ion and Two-Ion Anisotropies in Frustrated 2D.pdf;/Users/wasmer/Zotero/storage/FN7Y4K4H/htm.html}
}
@article{andersenOPTIMADEAPIExchanging2021,
@@ -84,28 +84,41 @@
url = {http://arxiv.org/abs/2103.02068},
urldate = {2021-10-15},
abstract = {The Open Databases Integration for Materials Design (OPTIMADE) consortium has designed a universal application programming interface (API) to make materials databases accessible and interoperable. We outline the first stable release of the specification, v1.0, which is already supported by many leading databases and several software packages. We illustrate the advantages of the OPTIMADE API through worked examples on each of the public materials databases that support the full API specification.},
- archiveprefix = {arXiv},
keywords = {Condensed Matter - Materials Science},
- file = {/home/johannes/Nextcloud/Zotero/Andersen et al_2021_OPTIMADE, an API for exchanging materials data.pdf;/home/johannes/Zotero/storage/PI7C4VKS/2103.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Andersen et al_2021_OPTIMADE, an API for exchanging materials data.pdf;/Users/wasmer/Zotero/storage/PI7C4VKS/2103.html}
}
-@misc{andersonCormorantCovariantMolecular2019,
+@online{andersonCormorantCovariantMolecular2019,
title = {Cormorant: {{Covariant Molecular Neural Networks}}},
shorttitle = {Cormorant},
author = {Anderson, Brandon and Hy, Truong-Son and Kondor, Risi},
date = {2019-11-25},
number = {arXiv:1906.04015},
- eprint = {1906.04015},
+ eprint = {arXiv:1906.04015},
eprinttype = {arxiv},
- primaryclass = {physics, stat},
- publisher = {{arXiv}},
doi = {10.48550/arXiv.1906.04015},
url = {http://arxiv.org/abs/1906.04015},
urldate = {2022-10-04},
abstract = {We propose Cormorant, a rotationally covariant neural network architecture for learning the behavior and properties of complex many-body physical systems. We apply these networks to molecular systems with two goals: learning atomic potential energy surfaces for use in Molecular Dynamics simulations, and learning ground state properties of molecules calculated by Density Functional Theory. Some of the key features of our network are that (a) each neuron explicitly corresponds to a subset of atoms; (b) the activation of each neuron is covariant to rotations, ensuring that overall the network is fully rotationally invariant. Furthermore, the non-linearity in our network is based upon tensor products and the Clebsch-Gordan decomposition, allowing the network to operate entirely in Fourier space. Cormorant significantly outperforms competing algorithms in learning molecular Potential Energy Surfaces from conformational geometries in the MD-17 dataset, and is competitive with other methods at learning geometric, energetic, electronic, and thermodynamic properties of molecules on the GDB-9 dataset.},
- archiveprefix = {arXiv},
+ pubstate = {preprint},
keywords = {equivariant,GNN,MD17,ML,MLP,MPNN,O(3),QM9,representation learning,SchNet,SO(3)},
- file = {/home/johannes/Nextcloud/Zotero/Anderson et al_2019_Cormorant.pdf;/home/johannes/Zotero/storage/RY359LWP/1906.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Anderson et al_2019_Cormorant.pdf;/Users/wasmer/Zotero/storage/RY359LWP/1906.html}
+}
+
+@online{angelopoulosPredictionPoweredInference2023,
+ title = {Prediction-{{Powered Inference}}},
+ author = {Angelopoulos, Anastasios N. and Bates, Stephen and Fannjiang, Clara and Jordan, Michael I. and Zrnic, Tijana},
+ date = {2023-02-16},
+ number = {arXiv:2301.09633},
+ eprint = {arXiv:2301.09633},
+ eprinttype = {arxiv},
+ doi = {10.48550/arXiv.2301.09633},
+ url = {http://arxiv.org/abs/2301.09633},
+ urldate = {2023-03-01},
+ abstract = {We introduce prediction-powered inference \$\textbackslash unicode\{x2013\}\$ a framework for performing valid statistical inference when an experimental data set is supplemented with predictions from a machine-learning system. Our framework yields provably valid conclusions without making any assumptions on the machine-learning algorithm that supplies the predictions. Higher accuracy of the predictions translates to smaller confidence intervals, permitting more powerful inference. Prediction-powered inference yields simple algorithms for computing valid confidence intervals for statistical objects such as means, quantiles, and linear and logistic regression coefficients. We demonstrate the benefits of prediction-powered inference with data sets from proteomics, genomics, electronic voting, remote sensing, census analysis, and ecology.},
+ pubstate = {preprint},
+ keywords = {/unread,AlphaFold,DeepMind,General ML,ML,RCPS,risk,uncertainty quantification,with-code},
+ file = {/Users/wasmer/Nextcloud/Zotero/Angelopoulos et al_2023_Prediction-Powered Inference.pdf;/Users/wasmer/Zotero/storage/VUQUZZ32/2301.html}
}
@article{artrithEfficientAccurateMachinelearning2017,
@@ -122,7 +135,7 @@
url = {https://link.aps.org/doi/10.1103/PhysRevB.96.014112},
urldate = {2021-10-18},
abstract = {Machine-learning potentials (MLPs) for atomistic simulations are a promising alternative to conventional classical potentials. Current approaches rely on descriptors of the local atomic environment with dimensions that increase quadratically with the number of chemical species. In this paper, we demonstrate that such a scaling can be avoided in practice. We show that a mathematically simple and computationally efficient descriptor with constant complexity is sufficient to represent transition-metal oxide compositions and biomolecules containing 11 chemical species with a precision of around 3 meV/atom. This insight removes a perceived bound on the utility of MLPs and paves the way to investigate the physics of previously inaccessible materials with more than ten chemical species.},
- file = {/home/johannes/Nextcloud/Zotero/Artrith et al_2017_Efficient and accurate machine-learning interpolation of atomic energies in.pdf;/home/johannes/Zotero/storage/77VRNTN7/Artrith et al. - 2017 - Efficient and accurate machine-learning interpolat.pdf;/home/johannes/Zotero/storage/RL7TSVEA/PhysRevB.96.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Artrith et al_2017_Efficient and accurate machine-learning interpolation of atomic energies in.pdf;/Users/wasmer/Zotero/storage/77VRNTN7/Artrith et al. - 2017 - Efficient and accurate machine-learning interpolat.pdf;/Users/wasmer/Zotero/storage/RL7TSVEA/PhysRevB.96.html}
}
@article{atzGeometricDeepLearning2021,
@@ -144,7 +157,7 @@
langid = {english},
keywords = {CNN,equivariant,GCN,GDL,GNN,invariance,molecules,MPNN,review,review-of-GDL},
annotation = {Primary\_atype: Reviews Subject\_term: Cheminformatics;Computational models;Computational science Subject\_term\_id: cheminformatics;computational-models;computational-science},
- file = {/home/johannes/Nextcloud/Zotero/Atz et al_2021_Geometric deep learning on molecular representations.pdf;/home/johannes/Zotero/storage/WJWQFR9K/s42256-021-00418-8.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Atz et al_2021_Geometric deep learning on molecular representations.pdf;/Users/wasmer/Zotero/storage/WJWQFR9K/s42256-021-00418-8.html}
}
@article{bacTopologicalResponseAnomalous2022,
@@ -165,7 +178,7 @@
issue = {1},
langid = {english},
keywords = {Ferromagnetism,Magnetic properties and materials,Topological insulators},
- file = {/home/johannes/Nextcloud/Zotero/Bac et al_2022_Topological response of the anomalous Hall effect in MnBi2Te4 due to magnetic.pdf;/home/johannes/Nextcloud/Zotero/Bac et al_2022_Topological response of the anomalous Hall effect in MnBi2Te4 due to magnetic2_supplementary.pdf;/home/johannes/Zotero/storage/E6I5UGGJ/s41535-022-00455-5.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Bac et al_2022_Topological response of the anomalous Hall effect in MnBi2Te4 due to magnetic.pdf;/Users/wasmer/Nextcloud/Zotero/Bac et al_2022_Topological response of the anomalous Hall effect in MnBi2Te4 due to magnetic2_supplementary.pdf;/Users/wasmer/Zotero/storage/E6I5UGGJ/s41535-022-00455-5.html}
}
@article{bankoFastTrackResearchData2020,
@@ -184,12 +197,33 @@
urldate = {2022-05-12},
abstract = {Research data management is a major necessity for the digital transformation in material science. Material science is multifaceted and experimental data, especially, is highly diverse. We demonstrate an adjustable approach to a group level data management based on a customizable document management software. Our solution is to continuously transform data management workflows from generalized to specialized data management. We start up fast with a relatively unregulated base setting and adapt continuously over the period of use to transform more and more data procedures into specialized data management workflows. By continuous adaptation and integration of analysis workflows and metadata schemes, the amount and the quality of the data improves. As an example of this process, in a period of 36 months, data on over 1800 samples, mainly materials libraries with hundreds of individual samples, were collected. The research data management system now contains over 1700 deposition processes and more than 4000 characterization documents. From initially mainly user-defined data input, an increased number of specialized data processing workflows was developed allowing the collection of more specialized, quality-assured data sets.},
keywords = {experimental science,RDM},
- file = {/home/johannes/Nextcloud/Zotero/Banko_Ludwig_2020_Fast-Track to Research Data Management in Experimental Material Science–Setting.pdf;/home/johannes/Zotero/storage/7HEKG4XK/acscombsci.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Banko_Ludwig_2020_Fast-Track to Research Data Management in Experimental Material Science–Setting.pdf;/Users/wasmer/Zotero/storage/7HEKG4XK/acscombsci.html}
+}
+
+@article{barrettAutoregressiveNeuralnetworkWavefunctions2022,
+ title = {Autoregressive Neural-Network Wavefunctions for Ab Initio Quantum Chemistry},
+ author = {Barrett, Thomas D. and Malyshev, Aleksei and Lvovsky, A. I.},
+ date = {2022-04},
+ journaltitle = {Nature Machine Intelligence},
+ shortjournal = {Nat Mach Intell},
+ volume = {4},
+ number = {4},
+ pages = {351--358},
+ publisher = {{Nature Publishing Group}},
+ issn = {2522-5839},
+ doi = {10.1038/s42256-022-00461-z},
+ url = {https://www.nature.com/articles/s42256-022-00461-z},
+ urldate = {2023-01-25},
+ abstract = {In recent years, neural-network quantum states have emerged as powerful tools for the study of quantum many-body systems. Electronic structure calculations are one such canonical many-body problem that have attracted sustained research efforts spanning multiple decades, whilst only recently being attempted with neural-network quantum states. However, the complex non-local interactions and high sample complexity are substantial challenges that call for bespoke solutions. Here, we parameterize the electronic wavefunction with an autoregressive neural network that permits highly efficient and scalable sampling, whilst also embedding physical priors reflecting the structure of molecular systems without sacrificing expressibility. This allows us to perform electronic structure calculations on molecules with up to 30 spin orbitals—at least an order of magnitude more Slater determinants than previous applications of conventional neural-network quantum states—and we find that our ansatz can outperform the de facto gold-standard coupled-cluster methods even in the presence of strong quantum correlations. With a highly expressive neural network for which sampling is no longer a computational bottleneck, we conclude that the barriers to further scaling are not associated with the wavefunction ansatz itself, but rather are inherent to any variational Monte Carlo approach.},
+ issue = {4},
+ langid = {english},
+ keywords = {/unread,autoregressive,CC,coupled cluster,FCI,full configuration interaction,ML-ESM,ML-QM,ML-QMBP,neural network quantum state,NN,NNQS,prediction of wavefunction,rec-by-hongbin,Slater-Jastrow},
+ file = {/Users/wasmer/Nextcloud/Zotero/Barrett et al_2022_Autoregressive neural-network wavefunctions for ab initio quantum chemistry.pdf}
}
@article{barthLocalExchangecorrelationPotential1972,
title = {A Local Exchange-Correlation Potential for the Spin Polarized Case. i},
- author = {von Barth, U. and Hedin, L.},
+ author = {family=Barth, given=U., prefix=von, useprefix=false and Hedin, L.},
date = {1972-07},
journaltitle = {Journal of Physics C: Solid State Physics},
shortjournal = {J. Phys. C: Solid State Phys.},
@@ -204,7 +238,7 @@
abstract = {The local density theory is developed by Hohenberg, Kohn and Sham is extended to the spin polarized case. A spin dependent one- electron potential pertinent to ground state properties is obtained from calculations of the total energy per electron made with a 'bubble' (or random phase) type of dielectric function. The potential is found to be well represented by an analytic expression corresponding to a shifted and rescaled spin dependent Slater potential. To test this potential the momentum dependent spin susceptibility of an electron gas is calculated. The results compare favourably with available information from other calculations and from experiment. The potential obtained in this paper should be useful for split band calculations of magnetic materials.},
langid = {english},
keywords = {DFT,LDA,LSDA,original publication,xc functional},
- file = {/home/johannes/Nextcloud/Zotero/Barth_Hedin_1972_A local exchange-correlation potential for the spin polarized case.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Barth_Hedin_1972_A local exchange-correlation potential for the spin polarized case.pdf}
}
@book{bartok-partayGaussianApproximationPotential2010,
@@ -239,7 +273,7 @@
urldate = {2021-07-06},
abstract = {We introduce a class of interatomic potential models that can be automatically generated from data consisting of the energies and forces experienced by atoms, as derived from quantum mechanical calculations. The models do not have a fixed functional form and hence are capable of modeling complex potential energy landscapes. They are systematically improvable with more data. We apply the method to bulk crystals, and test it by calculating properties at high temperatures. Using the interatomic potential to generate the long molecular dynamics trajectories required for such calculations saves orders of magnitude in computational cost.},
keywords = {GAP,ML,models,original publication},
- file = {/home/johannes/Nextcloud/Zotero/Bartók et al_2010_Gaussian Approximation Potentials.pdf;/home/johannes/Zotero/storage/DQIZDC4R/Bartók et al. - 2010 - Gaussian Approximation Potentials The Accuracy of.pdf;/home/johannes/Zotero/storage/QQUERR3G/PhysRevLett.104.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Bartók et al_2010_Gaussian Approximation Potentials.pdf;/Users/wasmer/Zotero/storage/DQIZDC4R/Bartók et al. - 2010 - Gaussian Approximation Potentials The Accuracy of.pdf;/Users/wasmer/Zotero/storage/QQUERR3G/PhysRevLett.104.html}
}
@unpublished{bartokGaussianApproximationPotentials2020,
@@ -249,13 +283,12 @@
date = {2020-02-05},
eprint = {1502.01366},
eprinttype = {arxiv},
- primaryclass = {cond-mat, physics:physics},
+ eprintclass = {cond-mat, physics:physics},
url = {http://arxiv.org/abs/1502.01366},
urldate = {2021-07-06},
abstract = {We present a swift walk-through of our recent work that uses machine learning to fit interatomic potentials based on quantum mechanical data. We describe our Gaussian Approximation Potentials (GAP) framework, discussing a variety of descriptors, how to train the model on total energies and derivatives and the simultaneous use of multiple models. We also show a small example using QUIP, the software sandbox implementation of GAP that is available for non-commercial use.},
- archiveprefix = {arXiv},
keywords = {_tablet,GAP,ML,models,QUIP,SOAP,tutorial},
- file = {/home/johannes/Nextcloud/Zotero/Bartók_Csányi_2020_Gaussian Approximation Potentials.pdf;/home/johannes/Zotero/storage/SBML3RKM/1502.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Bartók_Csányi_2020_Gaussian Approximation Potentials.pdf;/Users/wasmer/Zotero/storage/SBML3RKM/1502.html}
}
@article{bartokMachineLearningUnifies2017,
@@ -270,7 +303,7 @@
abstract = {Statistical learning based on a local representation of atomic structures provides a universal model of chemical stability.},
langid = {english},
keywords = {CCSD(T),coupled cluster,DFT,GAP,GGA,kernel methods,ML,molecules,silicon,SOAP,solids,surface physics},
- file = {/home/johannes/Nextcloud/Zotero/Bartók et al_2017_Machine learning unifies the modeling of materials and molecules.pdf;/home/johannes/Zotero/storage/DZL84DP7/sciadv.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Bartók et al_2017_Machine learning unifies the modeling of materials and molecules.pdf;/Users/wasmer/Zotero/storage/DZL84DP7/sciadv.html}
}
@article{bartokRepresentingChemicalEnvironments2013,
@@ -283,7 +316,7 @@
number = {18},
doi = {10.1103/PhysRevB.87.184115},
keywords = {descriptors,ML,original publication,SOAP},
- file = {/home/johannes/Nextcloud/Zotero/Bartók_2013_On representing chemical environments.pdf;/home/johannes/Zotero/storage/VRNA6FAC/PhysRevB.87.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Bartók_2013_On representing chemical environments.pdf;/Users/wasmer/Zotero/storage/VRNA6FAC/PhysRevB.87.html}
}
@article{basovPropertiesDemandQuantum2017,
@@ -304,7 +337,7 @@
issue = {11},
langid = {english},
annotation = {Bandiera\_abtest: a Cg\_type: Nature Research Journals Primary\_atype: Reviews Subject\_term: Electronic properties and materials;Phase transitions and critical phenomena Subject\_term\_id: electronic-properties-and-materials;phase-transitions-and-critical-phenomena},
- file = {/home/johannes/Nextcloud/Zotero/Basov et al_2017_Towards properties on demand in quantum materials.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Basov et al_2017_Towards properties on demand in quantum materials.pdf}
}
@unpublished{batatiaDesignSpaceEquivariant2022,
@@ -314,34 +347,31 @@
number = {arXiv:2205.06643},
eprint = {2205.06643},
eprinttype = {arxiv},
- primaryclass = {cond-mat, physics:physics, stat},
+ eprintclass = {cond-mat, physics:physics, stat},
publisher = {{arXiv}},
doi = {10.48550/arXiv.2205.06643},
url = {http://arxiv.org/abs/2205.06643},
urldate = {2022-05-21},
abstract = {The rapid progress of machine learning interatomic potentials over the past couple of years produced a number of new architectures. Particularly notable among these are the Atomic Cluster Expansion (ACE), which unified many of the earlier ideas around atom density-based descriptors, and Neural Equivariant Interatomic Potentials (NequIP), a message passing neural network with equivariant features that showed state of the art accuracy. In this work, we construct a mathematical framework that unifies these models: ACE is generalised so that it can be recast as one layer of a multi-layer architecture. From another point of view, the linearised version of NequIP is understood as a particular sparsification of a much larger polynomial model. Our framework also provides a practical tool for systematically probing different choices in the unified design space. We demonstrate this by an ablation study of NequIP via a set of experiments looking at in- and out-of-domain accuracy and smooth extrapolation very far from the training data, and shed some light on which design choices are critical for achieving high accuracy. Finally, we present BOTNet (Body-Ordered-Tensor-Network), a much-simplified version of NequIP, which has an interpretable architecture and maintains accuracy on benchmark datasets.},
- archiveprefix = {arXiv},
keywords = {_tablet,ACE,BOTNet,descriptors,equivariant,GNN,ML,MLP,MPNN,NequIP,NN,unified theory},
- file = {/home/johannes/Nextcloud/Zotero/Batatia et al_2022_The Design Space of E(3)-Equivariant Atom-Centered Interatomic Potentials.pdf;/home/johannes/Zotero/storage/2FLTPTA2/2205.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Batatia et al_2022_The Design Space of E(3)-Equivariant Atom-Centered Interatomic Potentials.pdf;/Users/wasmer/Zotero/storage/2FLTPTA2/2205.html}
}
-@misc{batatiaMACEHigherOrder2022,
+@online{batatiaMACEHigherOrder2022,
title = {{{MACE}}: {{Higher Order Equivariant Message Passing Neural Networks}} for {{Fast}} and {{Accurate Force Fields}}},
shorttitle = {{{MACE}}},
author = {Batatia, Ilyes and Kovács, Dávid Péter and Simm, Gregor N. C. and Ortner, Christoph and Csányi, Gábor},
date = {2022-06-15},
number = {arXiv:2206.07697},
- eprint = {2206.07697},
+ eprint = {arXiv:2206.07697},
eprinttype = {arxiv},
- primaryclass = {cond-mat, physics:physics, stat},
- publisher = {{arXiv}},
doi = {10.48550/arXiv.2206.07697},
url = {http://arxiv.org/abs/2206.07697},
urldate = {2022-09-25},
abstract = {Creating fast and accurate force fields is a long-standing challenge in computational chemistry and materials science. Recently, several equivariant message passing neural networks (MPNNs) have been shown to outperform models built using other approaches in terms of accuracy. However, most MPNNs suffer from high computational cost and poor scalability. We propose that these limitations arise because MPNNs only pass two-body messages leading to a direct relationship between the number of layers and the expressivity of the network. In this work, we introduce MACE, a new equivariant MPNN model that uses higher body order messages. In particular, we show that using four-body messages reduces the required number of message passing iterations to just \textbackslash emph\{two\}, resulting in a fast and highly parallelizable model, reaching or exceeding state-of-the-art accuracy on the rMD17, 3BPA, and AcAc benchmark tasks. We also demonstrate that using higher order messages leads to an improved steepness of the learning curves.},
- archiveprefix = {arXiv},
+ pubstate = {preprint},
keywords = {_tablet,ACE,chemical species scaling problem,descriptors,equivariant,library,MACE,ML,MLP,models,MPNN,Multi-ACE,NequIP,original publication,unified theory,with-code},
- file = {/home/johannes/Nextcloud/Zotero/Batatia et al_2022_MACE.pdf;/home/johannes/Zotero/storage/LDAKZMRF/2206.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Batatia et al_2022_MACE.pdf;/Users/wasmer/Zotero/storage/LDAKZMRF/2206.html}
}
@article{batraEmergingMaterialsIntelligence2020,
@@ -358,7 +388,7 @@
abstract = {The age of cognitive computing and artificial intelligence (AI) is just dawning. Inspired by its successes and promises, several AI ecosystems are blossoming, many of them within the domain of materials science and engineering. These materials intelligence ecosystems are being shaped by several independent developments. Machine learning (ML) algorithms and extant materials data are utilized to create surrogate models of materials properties and performance predictions. Materials data repositories, which fuel such surrogate model development, are mushrooming. Automated data and knowledge capture from the literature (to populate data repositories) using natural language processing approaches is being explored. The design of materials that meet target property requirements and of synthesis steps to create target materials appear to be within reach, either by closed-loop active-learning strategies or by inverting the prediction pipeline using advanced generative algorithms. AI and ML concepts are also transforming the computational and physical laboratory infrastructural landscapes used to create materials data in the first place. Surrogate models that can outstrip physics-based simulations (on which they are trained) by several orders of magnitude in speed while preserving accuracy are being actively developed. Automation, autonomy and guided high-throughput techniques are imparting enormous efficiencies and eliminating redundancies in materials synthesis and characterization. The integration of the various parts of the burgeoning ML landscape may lead to materials-savvy digital assistants and to a human–machine partnership that could enable dramatic efficiencies, accelerated discoveries and increased productivity. Here, we review these emergent materials intelligence ecosystems and discuss the imminent challenges and opportunities.},
langid = {english},
keywords = {materials informatics},
- file = {/home/johannes/Nextcloud/Zotero/Batra et al_2020_Emerging materials intelligence ecosystems propelled by machine learning.pdf;/home/johannes/Zotero/storage/A3A6TGKC/s41578-020-00255-y.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Batra et al_2020_Emerging materials intelligence ecosystems propelled by machine learning.pdf;/Users/wasmer/Zotero/storage/A3A6TGKC/s41578-020-00255-y.html}
}
@unpublished{batznerEquivariantGraphNeural2021,
@@ -367,14 +397,13 @@
date = {2021-12-16},
eprint = {2101.03164},
eprinttype = {arxiv},
- primaryclass = {cond-mat, physics:physics},
+ eprintclass = {cond-mat, physics:physics},
url = {http://arxiv.org/abs/2101.03164},
urldate = {2022-01-02},
abstract = {This work presents Neural Equivariant Interatomic Potentials (NequIP), an E(3)-equivariant neural network approach for learning interatomic potentials from ab-initio calculations for molecular dynamics simulations. While most contemporary symmetry-aware models use invariant convolutions and only act on scalars, NequIP employs E(3)-equivariant convolutions for interactions of geometric tensors, resulting in a more information-rich and faithful representation of atomic environments. The method achieves state-of-the-art accuracy on a challenging and diverse set of molecules and materials while exhibiting remarkable data efficiency. NequIP outperforms existing models with up to three orders of magnitude fewer training data, challenging the widely held belief that deep neural networks require massive training sets. The high data efficiency of the method allows for the construction of accurate potentials using high-order quantum chemical level of theory as reference and enables high-fidelity molecular dynamics simulations over long time scales.},
- archiveprefix = {arXiv},
version = {3},
keywords = {_tablet,GNN,MD,ML,MLP,molecules,MPNN,NequIP,Neural networks,Physics - Computational Physics,solids},
- file = {/home/johannes/Nextcloud/Zotero/Batzner et al_2021_E(3)-Equivariant Graph Neural Networks for Data-Efficient and Accurate.pdf;/home/johannes/Nextcloud/Zotero/false;/home/johannes/Zotero/storage/85ATGPNR/s41467-022-29939-5.html;/home/johannes/Zotero/storage/V4Y8BWNW/2101.html}
+ file = {/home/johannes/Nextcloud/Zotero/false;/Users/wasmer/Nextcloud/Zotero/Batzner et al_2021_E(3)-Equivariant Graph Neural Networks for Data-Efficient and Accurate.pdf;/Users/wasmer/Zotero/storage/85ATGPNR/s41467-022-29939-5.html;/Users/wasmer/Zotero/storage/V4Y8BWNW/2101.html}
}
@thesis{bauerDevelopmentRelativisticFullpotential2014,
@@ -389,7 +418,7 @@
isbn = {9783893369348},
langid = {english},
keywords = {_tablet,juKKR,KKR,KKRimp,PGI-1/IAS-1,thesis},
- file = {/home/johannes/Nextcloud/Zotero/Bauer_2014_Development of a relativistic full-potential first-principles multiple.pdf;/home/johannes/Zotero/storage/SYS2ZV93/151022.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Bauer_2014_Development of a relativistic full-potential first-principles multiple.pdf;/Users/wasmer/Zotero/storage/SYS2ZV93/151022.html}
}
@article{beckeDensityfunctionalTheoryVs2022,
@@ -409,7 +438,7 @@
urldate = {2022-12-29},
abstract = {In a recent paper [A. D. Becke, J. Chem. Phys. 156, 214101 (2022)], we compared two Kohn–Sham density functionals based on physical modeling and theory with the best density-functional power series fits in the literature. With only a handful of physically motivated pre-factors, our functionals matched, and even slightly exceeded, the performance of the best power-series functionals on the general main group thermochemistry, kinetics, and noncovalent interactions (GMTKN55) chemical database of Goerigk et al. [Phys. Chem. Chem. Phys. 19, 32184 (2017)]. This begs the question: how much can their performance be improved by adding power-series terms of our own? We address this question in the present work. First, we describe a series expansion variable that we believe contains more local physics than any other variable considered to date. Then we undertake modest, one-dimensional fits to the GMTKN55 data with our theory-based functional corrected by power-series exchange and dynamical correlation terms. We settle on 12 power-series terms (plus six parent terms) and achieve the lowest GMTKN55 “WTMAD2†error yet reported, by a substantial margin, for a hybrid Kohn–Sham density functional. The new functional is called “B22plus.â€},
keywords = {/unread,B22,B22plus,DFA,DFT,DM21},
- file = {/home/johannes/Nextcloud/Zotero/Becke_2022_Density-functional theory vs density-functional fits.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Becke_2022_Density-functional theory vs density-functional fits.pdf}
}
@article{behlerAtomcenteredSymmetryFunctions2011,
@@ -428,7 +457,7 @@
urldate = {2021-05-18},
abstract = {Neural networks offer an unbiased and numerically very accurate approach to represent high-dimensional ab initio potential-energy surfaces. Once constructed, neural network potentials can provide the energies and forces many orders of magnitude faster than electronic structure calculations, and thus enable molecular dynamics simulations of large systems. However, Cartesian coordinates are not a good choice to represent the atomic positions, and a transformation to symmetry functions is required. Using simple benchmark systems, the properties of several types of symmetry functions suitable for the construction of high-dimensional neural network potential-energy surfaces are discussed in detail. The symmetry functions are general and can be applied to all types of systems such as molecules, crystalline and amorphous solids, and liquids.},
keywords = {ACSF,descriptors,HDNNP,ML,models,original publication},
- file = {/home/johannes/Zotero/storage/HVL86IPH/1.html}
+ file = {/Users/wasmer/Zotero/storage/HVL86IPH/1.html}
}
@article{behlerConstructingHighdimensionalNeural2015,
@@ -447,8 +476,7 @@
abstract = {A lot of progress has been made in recent years in the development of atomistic potentials using machine learning (ML) techniques. In contrast to most conventional potentials, which are based on physical approximations and simplifications to derive an analytic functional relation between the atomic configuration and the potential-energy, ML potentials rely on simple but very flexible mathematical terms without a direct physical meaning. Instead, in case of ML potentials the topology of the potential-energy surface is “learned†by adjusting a number of parameters with the aim to reproduce a set of reference electronic structure data as accurately as possible. Due to this bias-free construction, they are applicable to a wide range of systems without changes in their functional form, and a very high accuracy close to the underlying first-principles data can be obtained. Neural network potentials (NNPs), which have first been proposed about two decades ago, are an important class of ML potentials. Although the first NNPs have been restricted to small molecules with only a few degrees of freedom, they are now applicable to high-dimensional systems containing thousands of atoms, which enables addressing a variety of problems in chemistry, physics, and materials science. In this tutorial review, the basic ideas of NNPs are presented with a special focus on developing NNPs for high-dimensional condensed systems. A recipe for the construction of these potentials is given and remaining limitations of the method are discussed. © 2015 Wiley Periodicals, Inc.},
langid = {english},
keywords = {_tablet,HDNNP,ML,models,molecular dynamics,neural network potentials,review,tutorial},
- annotation = {\_eprint: https://onlinelibrary.wiley.com/doi/pdf/10.1002/qua.24890},
- file = {/home/johannes/Nextcloud/Zotero/Behler_2015_Constructing high-dimensional neural network potentials.pdf;/home/johannes/Zotero/storage/DQEEE6BV/qua.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Behler_2015_Constructing high-dimensional neural network potentials.pdf;/Users/wasmer/Zotero/storage/DQEEE6BV/qua.html}
}
@article{behlerFourGenerationsHighDimensional2021,
@@ -464,7 +492,7 @@
urldate = {2021-05-18},
abstract = {Since their introduction about 25 years ago, machine learning (ML) potentials have become an important tool in the field of atomistic simulations. After the initial decade, in which neural networks were successfully used to construct potentials for rather small molecular systems, the development of high-dimensional neural network potentials (HDNNPs) in 2007 opened the way for the application of ML potentials in simulations of large systems containing thousands of atoms. To date, many other types of ML potentials have been proposed continuously increasing the range of problems that can be studied. In this review, the methodology of the family of HDNNPs including new recent developments will be discussed using a classification scheme into four generations of potentials, which is also applicable to many other types of ML potentials. The first generation is formed by early neural network potentials designed for low-dimensional systems. High-dimensional neural network potentials established the second generation and are based on three key steps: first, the expression of the total energy as a sum of environment-dependent atomic energy contributions; second, the description of the atomic environments by atom-centered symmetry functions as descriptors fulfilling the requirements of rotational, translational, and permutation invariance; and third, the iterative construction of the reference electronic structure data sets by active learning. In third-generation HDNNPs, in addition, long-range interactions are included employing environment-dependent partial charges expressed by atomic neural networks. In fourth-generation HDNNPs, which are just emerging, in addition, nonlocal phenomena such as long-range charge transfer can be included. The applicability and remaining limitations of HDNNPs are discussed along with an outlook at possible future developments.},
keywords = {HDNNP,ML,MLP,models,review,review-of-MLP},
- file = {/home/johannes/Nextcloud/Zotero/Behler_2021_Four Generations of High-Dimensional Neural Network Potentials.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Behler_2021_Four Generations of High-Dimensional Neural Network Potentials.pdf}
}
@article{behlerGeneralizedNeuralNetworkRepresentation2007,
@@ -477,7 +505,7 @@
number = {14},
doi = {10.1103/PhysRevLett.98.146401},
keywords = {BPNN,MD,ML,MLP,models,NN,original publication},
- file = {/home/johannes/Zotero/storage/RNTYUSXX/PhysRevLett.98.html}
+ file = {/Users/wasmer/Zotero/storage/RNTYUSXX/PhysRevLett.98.html}
}
@article{behlerPerspectiveMachineLearning2016,
@@ -497,7 +525,7 @@
urldate = {2021-05-18},
abstract = {Nowadays, computer simulations have become a standard tool in essentially all fields of chemistry, condensed matter physics, and materials science. In order to keep up with state-of-the-art experiments and the ever growing complexity of the investigated problems, there is a constantly increasing need for simulations of more realistic, i.e., larger, model systems with improved accuracy. In many cases, the availability of sufficiently efficient interatomic potentials providing reliable energies and forces has become a serious bottleneck for performing these simulations. To address this problem, currently a paradigm change is taking place in the development of interatomic potentials. Since the early days of computer simulations simplified potentials have been derived using physical approximations whenever the direct application of electronic structure methods has been too demanding. Recent advances in machine learning (ML) now offer an alternative approach for the representation of potential-energy surfaces by fitting large data sets from electronic structure calculations. In this perspective, the central ideas underlying these ML potentials, solved problems and remaining challenges are reviewed along with a discussion of their current applicability and limitations.},
keywords = {ML,MLP,models,review},
- file = {/home/johannes/Nextcloud/Zotero/Behler_2016_Perspective.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Behler_2016_Perspective.pdf}
}
@article{bengioRepresentationLearningReview2013,
@@ -514,7 +542,7 @@
abstract = {The success of machine learning algorithms generally depends on data representation, and we hypothesize that this is because different representations can entangle and hide more or less the different explanatory factors of variation behind the data. Although specific domain knowledge can be used to help design representations, learning with generic priors can also be used, and the quest for AI is motivating the design of more powerful representation-learning algorithms implementing such priors. This paper reviews recent work in the area of unsupervised feature learning and deep learning, covering advances in probabilistic models, autoencoders, manifold learning, and deep networks. This motivates longer term unanswered questions about the appropriate objectives for learning good representations, for computing representations (i.e., inference), and the geometrical connections between representation learning, density estimation, and manifold learning.},
eventtitle = {{{IEEE Transactions}} on {{Pattern Analysis}} and {{Machine Intelligence}}},
keywords = {Abstracts,autoencoder,Boltzmann machine,Deep learning,Feature extraction,feature learning,Learning systems,Machine learning,Manifolds,neural nets,Neural networks,representation learning,Speech recognition,unsupervised learning},
- file = {/home/johannes/Nextcloud/Zotero/Bengio et al_2013_Representation Learning.pdf;/home/johannes/Zotero/storage/PEAGSIHD/6472238.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Bengio et al_2013_Representation Learning.pdf;/Users/wasmer/Zotero/storage/PEAGSIHD/6472238.html}
}
@article{benmahmoudLearningElectronicDensity2020,
@@ -532,7 +560,7 @@
urldate = {2021-10-20},
abstract = {The electronic density of states (DOS) quantifies the distribution of the energy levels that can be occupied by electrons in a quasiparticle picture and is central to modern electronic structure theory. It also underpins the computation and interpretation of experimentally observable material properties such as optical absorption and electrical conductivity. We discuss the challenges inherent in the construction of a machine-learning (ML) framework aimed at predicting the DOS as a combination of local contributions that depend in turn on the geometric configuration of neighbors around each atom, using quasiparticle energy levels from density functional theory as training data. We present a challenging case study that includes configurations of silicon spanning a broad set of thermodynamic conditions, ranging from bulk structures to clusters and from semiconducting to metallic behavior. We compare different approaches to represent the DOS, and the accuracy of predicting quantities such as the Fermi level, the electron density at the Fermi level, or the band energy, either directly or as a side product of the evaluation of the DOS. We find that the performance of the model depends crucially on the resolution chosen to smooth the DOS and that there is a tradeoff to be made between the systematic error associated with the smoothing and the error in the ML model for a specific structure. We find however that the errors are not strongly correlated among similar structures, and so the average DOS over an ensemble of configurations is in very good agreement with the reference electronic structure calculations, despite the large nominal error on individual configurations. We demonstrate the usefulness of this approach by computing the density of states of a large amorphous silicon sample, for which it would be prohibitively expensive to compute the DOS by direct electronic structure calculations and show how the atom-centered decomposition of the DOS that is obtained through our model can be used to extract physical insights into the connections between structural and electronic features.},
keywords = {DFT,GPR,KPCovR,ML,ML-DFT,ML-ESM,nonscalar learning target,PCovR,prediction of LDOS,SOAP},
- file = {/home/johannes/Nextcloud/Zotero/Ben Mahmoud et al_2020_Learning the electronic density of states in condensed matter.pdf;/home/johannes/Zotero/storage/AG3V7VGZ/Ben Mahmoud et al_2020_Learning the electronic density of states in condensed matter2.pdf;/home/johannes/Zotero/storage/BIS7Q3X7/PhysRevB.102.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Ben Mahmoud et al_2020_Learning the electronic density of states in condensed matter.pdf;/Users/wasmer/Zotero/storage/AG3V7VGZ/Ben Mahmoud et al_2020_Learning the electronic density of states in condensed matter2.pdf;/Users/wasmer/Zotero/storage/BIS7Q3X7/PhysRevB.102.html}
}
@article{benmahmoudPredictingHotelectronFree2022,
@@ -550,7 +578,7 @@
urldate = {2022-09-28},
abstract = {Machine-learning potentials are usually trained on the ground-state, Born-Oppenheimer energy surface, which depends exclusively on the atomic positions and not on the simulation temperature. This disregards the effect of thermally excited electrons, that is important in metals, and essential to the description of warm dense matter. An accurate physical description of these effects requires that the nuclei move on a temperature-dependent electronic free energy. We propose a method to obtain machine-learning predictions of this free energy at an arbitrary electron temperature using exclusively training data from ground-state calculations, avoiding the need to train temperature-dependent potentials, and benchmark it on metallic liquid hydrogen at the conditions of the core of gas giants and brown dwarfs. This Letter demonstrates the advantages of hybrid schemes that use physical consideration to combine machine-learning predictions, providing a blueprint for the development of similar approaches that extend the reach of atomistic modeling by removing the barrier between physics and data-driven methodologies.},
keywords = {approximative GPR,EOS,extrapolate from gound state,finite-temperature DFT,forces,GAP,GPR,Hellmann-Feynman,ML,ML-DFT,ML-ESM,prediction from DOS,prediction of free energy,SOAP,warm dense matter},
- file = {/home/johannes/Nextcloud/Zotero/Ben Mahmoud et al_2022_Predicting hot-electron free energies from ground-state data.pdf;/home/johannes/Zotero/storage/6U9PWZG6/Ben Mahmoud et al. - 2022 - Predicting hot-electron free energies from ground-.pdf;/home/johannes/Zotero/storage/5YSTIB2N/PhysRevB.106.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Ben Mahmoud et al_2022_Predicting hot-electron free energies from ground-state data.pdf;/Users/wasmer/Zotero/storage/6U9PWZG6/Ben Mahmoud et al. - 2022 - Predicting hot-electron free energies from ground-.pdf;/Users/wasmer/Zotero/storage/5YSTIB2N/PhysRevB.106.html}
}
@unpublished{bernerModernMathematicsDeep2021,
@@ -559,13 +587,12 @@
date = {2021-05-09},
eprint = {2105.04026},
eprinttype = {arxiv},
- primaryclass = {cs, stat},
+ eprintclass = {cs, stat},
url = {http://arxiv.org/abs/2105.04026},
urldate = {2022-01-02},
abstract = {We describe the new field of mathematical analysis of deep learning. This field emerged around a list of research questions that were not answered within the classical framework of learning theory. These questions concern: the outstanding generalization power of overparametrized neural networks, the role of depth in deep architectures, the apparent absence of the curse of dimensionality, the surprisingly successful optimization performance despite the non-convexity of the problem, understanding what features are learned, why deep architectures perform exceptionally well in physical problems, and which fine aspects of an architecture affect the behavior of a learning task in which way. We present an overview of modern approaches that yield partial answers to these questions. For selected approaches, we describe the main ideas in more detail.},
- archiveprefix = {arXiv},
keywords = {Computer Science - Machine Learning,Statistics - Machine Learning},
- file = {/home/johannes/Nextcloud/Zotero/Berner et al_2021_The Modern Mathematics of Deep Learning.pdf;/home/johannes/Zotero/storage/XDBSS3FE/2105.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Berner et al_2021_The Modern Mathematics of Deep Learning.pdf;/Users/wasmer/Zotero/storage/XDBSS3FE/2105.html}
}
@article{bigiSmoothBasisAtomistic2022,
@@ -583,8 +610,23 @@
url = {https://aip.scitation.org/doi/10.1063/5.0124363},
urldate = {2022-12-29},
abstract = {Machine learning frameworks based on correlations of interatomic positions begin with a discretized description of the density of other atoms in the neighborhood of each atom in the system. Symmetry considerations support the use of spherical harmonics to expand the angular dependence of this density, but there is, as of yet, no clear rationale to choose one radial basis over another. Here, we investigate the basis that results from the solution of the Laplacian eigenvalue problem within a sphere around the atom of interest. We show that this generates a basis of controllable smoothness within the sphere (in the same sense as plane waves provide a basis with controllable smoothness for a problem with periodic boundaries) and that a tensor product of Laplacian eigenstates also provides a smooth basis for expanding any higher-order correlation of the atomic density within the appropriate hypersphere. We consider several unsupervised metrics of the quality of a basis for a given dataset and show that the Laplacian eigenstate basis has a performance that is much better than some widely used basis sets and competitive with data-driven bases that numerically optimize each metric. Finally, we investigate the role of the basis in building models of the potential energy. In these tests, we find that a combination of the Laplacian eigenstate basis and target-oriented heuristics leads to equal or improved regression performance when compared to both heuristic and data-driven bases in the literature. We conclude that the smoothness of the basis functions is a key aspect of successful atomic density representations.},
- keywords = {/unread,ACDC,ACE,density correlation,descriptor comparison,descriptors,equivariant,general body-order,GPR,Jacobian condition number,Laplacian eigenstate basis,MACE,NN,prediction of potential energy,radial basis,regression,residual variance,smooth basis,Supervised learning,unsupervised learning},
- file = {/home/johannes/Nextcloud/Zotero/Bigi et al_2022_A smooth basis for atomistic machine learning.pdf;/home/johannes/Nextcloud/Zotero/Bigi et al_2022_A smooth basis for atomistic machine learning2.pdf}
+ keywords = {_tablet,/unread,ACDC,ACE,density correlation,descriptor comparison,descriptors,equivariant,general body-order,GPR,Jacobian condition number,Laplacian eigenstate basis,MACE,NN,prediction of potential energy,radial basis,regression,residual variance,smooth basis,Supervised learning,unsupervised learning},
+ file = {/home/johannes/Nextcloud/Zotero/false;/Users/wasmer/Nextcloud/Zotero/Bigi et al_2022_A smooth basis for atomistic machine learning.pdf}
+}
+
+@online{bishopPlenaryFifthParadigm2022,
+ type = {Video},
+ title = {Plenary: {{The}} Fifth Paradigm of Scientific Discovery},
+ shorttitle = {Plenary},
+ author = {Bishop, Christopher Michael and Welling, Max and LLorens, Ashley},
+ date = {2022-10-17},
+ url = {https://www.microsoft.com/en-us/research/video/plenary-the-fifth-paradigm-of-scientific-discovery/},
+ urldate = {2023-01-16},
+ abstract = {Chris Bishop, technical fellow and director of Microsoft Research AI4Science joins colleagues and collaborators across Microsoft Research to discuss how deep learning is set to have a transformational impact on the sciences – including potential applications for drug discovery and materials design. Learn more about the 2022 Microsoft Research Summit […]},
+ langid = {american},
+ organization = {{Microsoft Research Summit 2022}},
+ keywords = {/unread},
+ file = {/Users/wasmer/Zotero/storage/IJ8MX5EV/plenary-the-fifth-paradigm-of-scientific-discovery.html}
}
@software{blaiszikChartingMLPublications2022,
@@ -611,7 +653,7 @@
doi = {10.1063/1.469597},
url = {https://aip.scitation.org/doi/10.1063/1.469597},
urldate = {2021-10-22},
- file = {/home/johannes/Nextcloud/Zotero/Blank et al_1995_Neural network models of potential energy surfaces.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Blank et al_1995_Neural network models of potential energy surfaces.pdf}
}
@report{bluegelComputationalCondensedMatter2006,
@@ -627,7 +669,7 @@
isbn = {9783893364305},
langid = {ngerman},
keywords = {condensed matter,DFT,FZJ,IFF,IFF spring school,KKR,magnetism,PGI-1/IAS-1},
- file = {/home/johannes/Nextcloud/Zotero/Blügel et al_2006_Computational Condensed Matter Physics.pdf;/home/johannes/Zotero/storage/IUT3QPKV/56047.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Blügel et al_2006_Computational Condensed Matter Physics.pdf;/Users/wasmer/Zotero/storage/IUT3QPKV/56047.html}
}
@report{blugelDensityFunctionalTheory2006,
@@ -642,7 +684,7 @@
isbn = {9783893364305},
langid = {english},
keywords = {_tablet,bluegel,DFT,FLEUR,IFF,IFF spring school,PGI-1/IAS-1,rec-by-bluegel},
- file = {/home/johannes/Nextcloud/Zotero/Blügel_2006_Density Functional Theory in Practice.pdf;/home/johannes/Zotero/storage/ZL4WZAY7/51316.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Blügel_2006_Density Functional Theory in Practice.pdf;/Users/wasmer/Zotero/storage/ZL4WZAY7/51316.html}
}
@book{blumFoundationsDataScience2020,
@@ -672,7 +714,7 @@
langid = {english},
keywords = {condensed matter,magnetism,textbook,undergraduate},
annotation = {OCLC: 53956469},
- file = {/home/johannes/Nextcloud/Zotero/Blundell_2001_Magnetism in condensed matter.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Blundell_2001_Magnetism in condensed matter.pdf}
}
@article{bochkarevEfficientParametrizationAtomic2022,
@@ -685,7 +727,7 @@
number = {1},
doi = {10.1103/PhysRevMaterials.6.013804},
keywords = {_tablet,ACE,descriptors,library,ML,pacemaker,with-code},
- file = {/home/johannes/Nextcloud/Zotero/Bochkarev_2022_Efficient parametrization of the atomic cluster expansion.pdf;/home/johannes/Zotero/storage/LLPTMRGA/PhysRevMaterials.6.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Bochkarev_2022_Efficient parametrization of the atomic cluster expansion.pdf;/Users/wasmer/Zotero/storage/LLPTMRGA/PhysRevMaterials.6.html}
}
@unpublished{bochkarevMultilayerAtomicCluster2022,
@@ -695,33 +737,30 @@
number = {arXiv:2205.08177},
eprint = {2205.08177},
eprinttype = {arxiv},
- primaryclass = {cond-mat},
+ eprintclass = {cond-mat},
publisher = {{arXiv}},
doi = {10.48550/arXiv.2205.08177},
url = {http://arxiv.org/abs/2205.08177},
urldate = {2022-05-21},
abstract = {Traditionally, interatomic potentials assume local bond formation supplemented by long-range electrostatic interactions when necessary. This ignores intermediate range multi-atom interactions that arise from the relaxation of the electronic structure. Here, we present the multilayer atomic cluster expansion (ml-ACE) that includes collective, semi-local multi-atom interactions naturally within its remit. We demonstrate that ml-ACE significantly improves fit accuracy compared to a local expansion on selected examples and provide physical intuition to understand this improvement.},
- archiveprefix = {arXiv},
keywords = {_tablet,ACE,descriptors,ML,ml-ACE},
- file = {/home/johannes/Nextcloud/Zotero/Bochkarev et al_2022_Multilayer atomic cluster expansion for semi-local interactions.pdf;/home/johannes/Zotero/storage/NQ2MH8V7/2205.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Bochkarev et al_2022_Multilayer atomic cluster expansion for semi-local interactions.pdf;/Users/wasmer/Zotero/storage/NQ2MH8V7/2205.html}
}
-@misc{bochkarevMultilayerAtomicCluster2022a,
+@online{bochkarevMultilayerAtomicCluster2022a,
title = {Multilayer Atomic Cluster Expansion for Semi-Local Interactions},
author = {Bochkarev, Anton and Lysogorskiy, Yury and Ortner, Christoph and Csányi, Gábor and Drautz, Ralf},
date = {2022-05-17},
number = {arXiv:2205.08177},
- eprint = {2205.08177},
+ eprint = {arXiv:2205.08177},
eprinttype = {arxiv},
- primaryclass = {cond-mat},
- publisher = {{arXiv}},
doi = {10.48550/arXiv.2205.08177},
url = {http://arxiv.org/abs/2205.08177},
urldate = {2022-09-29},
abstract = {Traditionally, interatomic potentials assume local bond formation supplemented by long-range electrostatic interactions when necessary. This ignores intermediate range multi-atom interactions that arise from the relaxation of the electronic structure. Here, we present the multilayer atomic cluster expansion (ml-ACE) that includes collective, semi-local multi-atom interactions naturally within its remit. We demonstrate that ml-ACE significantly improves fit accuracy compared to a local expansion on selected examples and provide physical intuition to understand this improvement.},
- archiveprefix = {arXiv},
+ pubstate = {preprint},
keywords = {ACE,descriptors,ML,MLP,multilayer-ACE,semilocal interactions},
- file = {/home/johannes/Nextcloud/Zotero/Bochkarev et al_2022_Multilayer atomic cluster expansion for semi-local interactions2.pdf;/home/johannes/Zotero/storage/ZVU3IARD/2205.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Bochkarev et al_2022_Multilayer atomic cluster expansion for semi-local interactions2.pdf;/Users/wasmer/Zotero/storage/ZVU3IARD/2205.html}
}
@article{bockHybridModellingMachine2021,
@@ -741,25 +780,23 @@
issue = {8},
langid = {english},
keywords = {ANN,feature engineering,FEM,physics-informed ML},
- file = {/home/johannes/Nextcloud/Zotero/Bock et al_2021_Hybrid Modelling by Machine Learning Corrections of Analytical Model.pdf;/home/johannes/Zotero/storage/IN7CCMRJ/htm.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Bock et al_2021_Hybrid Modelling by Machine Learning Corrections of Analytical Model.pdf;/Users/wasmer/Zotero/storage/IN7CCMRJ/htm.html}
}
-@misc{bogojeskiEfficientPrediction3D2018,
+@online{bogojeskiEfficientPrediction3D2018,
title = {Efficient Prediction of {{3D}} Electron Densities Using Machine Learning},
author = {Bogojeski, Mihail and Brockherde, Felix and Vogt-Maranto, Leslie and Li, Li and Tuckerman, Mark E. and Burke, Kieron and Müller, Klaus-Robert},
date = {2018-11-15},
number = {arXiv:1811.06255},
- eprint = {1811.06255},
+ eprint = {arXiv:1811.06255},
eprinttype = {arxiv},
- primaryclass = {physics},
- publisher = {{arXiv}},
doi = {10.48550/arXiv.1811.06255},
url = {http://arxiv.org/abs/1811.06255},
urldate = {2022-07-08},
abstract = {The Kohn-Sham scheme of density functional theory is one of the most widely used methods to solve electronic structure problems for a vast variety of atomistic systems across different scientific fields. While the method is fast relative to other first principles methods and widely successful, the computational time needed is still not negligible, making it difficult to perform calculations for very large systems or over long time-scales. In this submission, we revisit a machine learning model capable of learning the electron density and the corresponding energy functional based on a set of training examples. It allows us to bypass solving the Kohn-Sham equations, providing a significant decrease in computation time. We specifically focus on the machine learning formulation of the Hohenberg-Kohn map and its decomposability. We give results and discuss challenges, limits and future directions.},
- archiveprefix = {arXiv},
+ pubstate = {preprint},
keywords = {_tablet,DFT,HK map,ML,ML-DFT,ML-ESM,ML-HK map,molecules,prediction of electron density},
- file = {/home/johannes/Nextcloud/Zotero/Bogojeski et al_2018_Efficient prediction of 3D electron densities using machine learning.pdf;/home/johannes/Zotero/storage/MCBT39D4/1811.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Bogojeski et al_2018_Efficient prediction of 3D electron densities using machine learning.pdf;/Users/wasmer/Zotero/storage/MCBT39D4/1811.html}
}
@article{bogojeskiQuantumChemicalAccuracy2020,
@@ -781,7 +818,7 @@
langid = {english},
keywords = {_tablet,2-step model,CCSD(T),coupled cluster,Delta,delta learning,DFT,HK map,KKR,ML,ML-DFA,ML-DFT,ML-ESM,ML-HK map,molecules,prediction of electron density,with-code,Δ-machine learning},
annotation = {Bandiera\_abtest: a Cc\_license\_type: cc\_by Cg\_type: Nature Research Journals Primary\_atype: Research Subject\_term: Computational chemistry;Computational science Subject\_term\_id: computational-chemistry;computational-science},
- file = {/home/johannes/Nextcloud/Zotero/Bogojeski et al_2020_Quantum chemical accuracy from density functional approximations via machine.pdf;/home/johannes/Nextcloud/Zotero/false}
+ file = {/home/johannes/Nextcloud/Zotero/false;/Users/wasmer/Nextcloud/Zotero/Bogojeski et al_2020_Quantum chemical accuracy from density functional approximations via machine.pdf}
}
@article{borchaniSurveyMultioutputRegression2015,
@@ -799,8 +836,7 @@
abstract = {In recent years, a plethora of approaches have been proposed to deal with the increasingly challenging task of multi-output regression. This study provides a survey on state-of-the-art multi-output regression methods, that are categorized as problem transformation and algorithm adaptation methods. In addition, we present the mostly used performance evaluation measures, publicly available data sets for multi-output regression real-world problems, as well as open-source software frameworks. WIREs Data Mining Knowl Discov 2015, 5:216–233. doi: 10.1002/widm.1157 This article is categorized under: Technologies {$>$} Machine Learning},
langid = {english},
keywords = {ML,multi-output learning,multi-target learning,regression},
- annotation = {\_eprint: https://wires.onlinelibrary.wiley.com/doi/pdf/10.1002/widm.1157},
- file = {/home/johannes/Nextcloud/Zotero/Borchani et al_2015_A survey on multi-output regression.pdf;/home/johannes/Zotero/storage/FKMFKWW3/widm.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Borchani et al_2015_A survey on multi-output regression.pdf;/Users/wasmer/Zotero/storage/FKMFKWW3/widm.html}
}
@thesis{bornemannLargescaleInvestigationsNontrivial2019,
@@ -815,7 +851,7 @@
isbn = {9783958063945},
langid = {english},
keywords = {juKKR,KKR,PGI-1/IAS-1,thesis},
- file = {/home/johannes/Nextcloud/Zotero/Bornemann_2019_Large-scale Investigations of Non-trivial Magnetic Textures in Chiral Magnets.pdf;/home/johannes/Zotero/storage/BZP7D4IW/861845.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Bornemann_2019_Large-scale Investigations of Non-trivial Magnetic Textures in Chiral Magnets.pdf;/Users/wasmer/Zotero/storage/BZP7D4IW/861845.html}
}
@article{bouazizSpinDynamics3d2019,
@@ -829,7 +865,7 @@
number = {5},
doi = {10.1103/PhysRevMaterials.3.054201},
keywords = {_tablet,defects,Funsilab,impurity embedding,PGI-1/IAS-1,topological insulator},
- file = {/home/johannes/Nextcloud/Zotero/Bouaziz_2019_Spin dynamics of 3d and 4d impurities embedded in prototypical topological.pdf;/home/johannes/Zotero/storage/CW3GMSS2/PhysRevMaterials.3.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Bouaziz_2019_Spin dynamics of 3d and 4d impurities embedded in prototypical topological.pdf;/Users/wasmer/Zotero/storage/CW3GMSS2/PhysRevMaterials.3.html}
}
@thesis{bouazizSpinorbitronicsNanoscaleAnalytical2019,
@@ -845,7 +881,7 @@
isbn = {9783958064294},
langid = {english},
keywords = {Hall QHE,Hall QSHE,juKKR,KKR,PGI-1/IAS-1,skyrmions,thesis,topological insulator},
- file = {/home/johannes/Nextcloud/Zotero/Bouaziz_2019_Spin-orbitronics at the nanoscale.pdf;/home/johannes/Zotero/storage/YM28TKHA/865993.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Bouaziz_2019_Spin-orbitronics at the nanoscale.pdf;/Users/wasmer/Zotero/storage/YM28TKHA/865993.html}
}
@article{brackTenSimpleRules2022,
@@ -864,7 +900,7 @@
urldate = {2022-07-28},
langid = {english},
keywords = {best practices,RSE,scientific workflows},
- file = {/home/johannes/Nextcloud/Zotero/Brack et al_2022_Ten simple rules for making a software tool workflow-ready.pdf;/home/johannes/Zotero/storage/QPHMATCQ/article.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Brack et al_2022_Ten simple rules for making a software tool workflow-ready.pdf;/Users/wasmer/Zotero/storage/QPHMATCQ/article.html}
}
@unpublished{brandstetterLiePointSymmetry2022,
@@ -874,15 +910,14 @@
number = {arXiv:2202.07643},
eprint = {2202.07643},
eprinttype = {arxiv},
- primaryclass = {cs},
+ eprintclass = {cs},
publisher = {{arXiv}},
doi = {10.48550/arXiv.2202.07643},
url = {http://arxiv.org/abs/2202.07643},
urldate = {2022-06-09},
abstract = {Neural networks are increasingly being used to solve partial differential equations (PDEs), replacing slower numerical solvers. However, a critical issue is that neural PDE solvers require high-quality ground truth data, which usually must come from the very solvers they are designed to replace. Thus, we are presented with a proverbial chicken-and-egg problem. In this paper, we present a method, which can partially alleviate this problem, by improving neural PDE solver sample complexity -- Lie point symmetry data augmentation (LPSDA). In the context of PDEs, it turns out that we are able to quantitatively derive an exhaustive list of data transformations, based on the Lie point symmetry group of the PDEs in question, something not possible in other application areas. We present this framework and demonstrate how it can easily be deployed to improve neural PDE solver sample complexity by an order of magnitude.},
- archiveprefix = {arXiv},
keywords = {data augmentation,ML,neural PDE solver,PDE,PINN,symmetry,with-code},
- file = {/home/johannes/Nextcloud/Zotero/Brandstetter et al_2022_Lie Point Symmetry Data Augmentation for Neural PDE Solvers.pdf;/home/johannes/Zotero/storage/QUUR7MZV/2202.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Brandstetter et al_2022_Lie Point Symmetry Data Augmentation for Neural PDE Solvers.pdf;/Users/wasmer/Zotero/storage/QUUR7MZV/2202.html}
}
@article{brandtKadi4MatResearchData2021,
@@ -903,7 +938,7 @@
issue = {1},
langid = {english},
keywords = {ELN,RDM,repository,RSE,workflows},
- file = {/home/johannes/Nextcloud/Zotero/Brandt et al_2021_Kadi4Mat.pdf;/home/johannes/Zotero/storage/GMTCYYBY/dsj-2021-008.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Brandt et al_2021_Kadi4Mat.pdf;/Users/wasmer/Zotero/storage/GMTCYYBY/dsj-2021-008.html}
}
@article{braunImpactSpinOrbit2021,
@@ -921,8 +956,7 @@
abstract = {Due to many important technical developments over the past two decades angle-resolved (inverse) photoemission has become the method of choice to study experimentally the bulk and surface-related electronic states of solids in the most detailed way. Due to new powerful photon sources as well as efficient analyzers and detectors extremely high energy and angle resolution are achieved nowadays for spin-integrated and also for spin-resolved measurements. These developments allow in particular to explore the influence of spin–orbit coupling on image potential states of simple metals like Ir, Pt, or Au with a high atomic number as well as new types of materials as for example topological insulators. Herein, fully relativistic angle- and spin-resolved inverse photoemission calculations are presented that make use of the spin-density matrix formulation of the one-step model. This way a quantitative analysis of all occupied and unoccupied electronic features in the vicinity of the Fermi level is achieved for a wide range of excitation energies. Using this approach, in addition, it is possible to deal with arbitrarily ordered but also disordered systems. Because of these features, the one-step or spectral function approach to photoemission permits detailed theoretical studies on a large variety of interesting solid-state systems.},
langid = {english},
keywords = {image states,photoemission,spin–orbit interaction},
- annotation = {\_eprint: https://onlinelibrary.wiley.com/doi/pdf/10.1002/pssb.202000026},
- file = {/home/johannes/Nextcloud/Zotero/Braun_Ebert_2021_The Impact of Spin–Orbit Interaction on the Image States of High-Z Materials.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Braun_Ebert_2021_The Impact of Spin–Orbit Interaction on the Image States of High-Z Materials.pdf}
}
@article{brockherdeBypassingKohnShamEquations2017,
@@ -943,7 +977,22 @@
issue = {1},
langid = {english},
keywords = {_tablet,DFT,HK map,KRR,ML,ML-DFT,ML-ESM,ML-HK map,ML-KS,ML-OF,prediction from potential,prediction of electron density},
- file = {/home/johannes/Nextcloud/Zotero/Brockherde et al_2017_Bypassing the Kohn-Sham equations with machine learning.pdf;/home/johannes/Zotero/storage/8X4ALINZ/s41467-017-00839-3.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Brockherde et al_2017_Bypassing the Kohn-Sham equations with machine learning.pdf;/Users/wasmer/Zotero/storage/8X4ALINZ/s41467-017-00839-3.html}
+}
+
+@book{broderHighthroughputAllElectronDensity2021,
+ title = {High-Throughput {{All-Electron Density Functional Theory Simulations}} for a {{Data-driven Chemical Interpretation}} of {{X-ray Photoelectron Spectra}}},
+ author = {Bröder, Jens},
+ date = {2021},
+ series = {Schriften Des {{Forschungszentrums Jülich}}. {{Reihe Schlüsseltechnologien}} / {{Key Technologies}}},
+ number = {229},
+ publisher = {{Forschungszentrum Jülich GmbH Zentralbibliothek, Verlag}},
+ location = {{Jülich}},
+ abstract = {Enabling computer-driven materials design to find and create materials with advanced propertiesfromthe enormous haystack of material phase space is a worthy goal for humanity. Most high-technologies, for example in the energy or health sector, strongly depend on advanced tailored materials. Since conventional research and screening of materials is rather slow and expensive, being able to determine material properties on the computer poses a paradigm shift. For the calculation of properties for pure materials on the nano scale ab initio methods based on the theory of quantum mechanics are well established. Density Functional Theory(DFT) is such a widely applied method from first principles with high predictive power. To screen through larger sets of atomic configurations physical property calculation processes need to be robust and automated. Automation is achieved through the deployment of advanced frameworks which manage many workflows while tracking the provenance of data and calculations. Through workflows, which are essential property calculator procedures, a high-level automation environment is achievable and accumulated knowledge can be reused by others. Workflows can be complex and include multiple programs solving problems over several physical length scales. In this work, the open source all-electron DFT program FLEUR implementing the highly accurate Full-potential Linearized Augmented Plane Wave (FLAPW) method is connected and deployed through the open source Automated Interactive Infrastructure and Database for Computational Science (AiiDA) framework to achieve automation. AiiDA is a Python framework which is capable of provenance tracking millions of high-through put simulations and their data. Basic and advanced workflows are implemented in an open source Pythonpackage AiiDA-FLEUR, especially to calculate properties for the chemical analysis of X-rayphotoemission spectra. These workflows are applied on a wide range of materials, in particular on most known metallic binary compounds. The chemical-phase composition and other material properties of a surface region can be understood through the careful chemical analysis of high-resolution X-ray photoemission spectra. The spectra evaluation process is improved through the development of a fittingmethod driven by data from ab initio simulations. For complex multi-phase spectra this proposedevaluation process is expected to have advantages over the widely applied conventional methods. The spectra evaluation process is successfully deployed on well-behaved spectra of materials relevant for the inner wall (blanket and divertor) plasma-facing components of a nuclear fusion reactor. In particular, the binary beryllium systems Be-Ti, Be-Wand Be-Ta are investigated. Furthermore, different approaches to calculate spectral properties like chemical shifts and binding energies are studied and benchmarked against the experimental literature and data from the NIST X-ray photoelectron spectroscopy database},
+ isbn = {978-3-95806-526-0},
+ langid = {english},
+ pagetotal = {169},
+ keywords = {/unread,AiiDA,AiiDA-FLEUR,DFT,FLEUR,high-throughput computing,PGI-1/IAS-1,thesis}
}
@unpublished{bronsteinGeometricDeepLearning2021,
@@ -953,13 +1002,12 @@
date = {2021-05-02},
eprint = {2104.13478},
eprinttype = {arxiv},
- primaryclass = {cs, stat},
+ eprintclass = {cs, stat},
url = {http://arxiv.org/abs/2104.13478},
urldate = {2022-04-14},
abstract = {The last decade has witnessed an experimental revolution in data science and machine learning, epitomised by deep learning methods. Indeed, many high-dimensional learning tasks previously thought to be beyond reach -- such as computer vision, playing Go, or protein folding -- are in fact feasible with appropriate computational scale. Remarkably, the essence of deep learning is built from two simple algorithmic principles: first, the notion of representation or feature learning, whereby adapted, often hierarchical, features capture the appropriate notion of regularity for each task, and second, learning by local gradient-descent type methods, typically implemented as backpropagation. While learning generic functions in high dimensions is a cursed estimation problem, most tasks of interest are not generic, and come with essential pre-defined regularities arising from the underlying low-dimensionality and structure of the physical world. This text is concerned with exposing these regularities through unified geometric principles that can be applied throughout a wide spectrum of applications. Such a 'geometric unification' endeavour, in the spirit of Felix Klein's Erlangen Program, serves a dual purpose: on one hand, it provides a common mathematical framework to study the most successful neural network architectures, such as CNNs, RNNs, GNNs, and Transformers. On the other hand, it gives a constructive procedure to incorporate prior physical knowledge into neural architectures and provide principled way to build future architectures yet to be invented.},
- archiveprefix = {arXiv},
keywords = {basics,CNN,Deep learning,equivariant,GCN,GDL,General ML,geometric deep learning,GNN,invariance,ML,MPNN,review,review-of-GDL,theory,tutorial},
- file = {/home/johannes/Nextcloud/Zotero/Bronstein et al_2021_Geometric Deep Learning.pdf;/home/johannes/Zotero/storage/6ZLIPHI5/2104.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Bronstein et al_2021_Geometric Deep Learning.pdf;/Users/wasmer/Zotero/storage/6ZLIPHI5/2104.html}
}
@article{burkeDFTNutshell2013,
@@ -977,27 +1025,24 @@
abstract = {The purpose of this short essay is to introduce students and other newcomers to the basic ideas and uses of modern electronic density functional theory, including what kinds of approximations are in current use, and how well they work (or not). The complete newcomer should find it orients them well, while even longtime users and aficionados might find something new outside their area. Important questions varying in difficulty and effort are posed in the text, and are answered in the Supporting Information. © 2012 Wiley Periodicals, Inc.},
langid = {english},
keywords = {_tablet,density functional theory,electronic structure,local density approximation},
- annotation = {\_eprint: https://onlinelibrary.wiley.com/doi/pdf/10.1002/qua.24259},
- file = {/home/johannes/Nextcloud/Zotero/Burke_Wagner_2013_DFT in a nutshell.pdf;/home/johannes/Nextcloud/Zotero/false;/home/johannes/Zotero/storage/CCPHAAVK/qua.html}
+ file = {/home/johannes/Nextcloud/Zotero/false;/Users/wasmer/Nextcloud/Zotero/Burke_Wagner_2013_DFT in a nutshell.pdf;/Users/wasmer/Zotero/storage/CCPHAAVK/qua.html}
}
-@misc{burkeLiesMyTeacher2021,
+@online{burkeLiesMyTeacher2021,
title = {Lies {{My Teacher Told Me About Density Functional Theory}}: {{Seeing Through Them}} with the {{Hubbard Dimer}}},
shorttitle = {Lies {{My Teacher Told Me About Density Functional Theory}}},
author = {Burke, Kieron and Kozlowski, John},
date = {2021-10-18},
number = {arXiv:2108.11534},
- eprint = {2108.11534},
+ eprint = {arXiv:2108.11534},
eprinttype = {arxiv},
- primaryclass = {cond-mat, physics:physics},
- publisher = {{arXiv}},
doi = {10.48550/arXiv.2108.11534},
url = {http://arxiv.org/abs/2108.11534},
urldate = {2022-07-10},
abstract = {Most realistic calculations of moderately correlated materials begin with a ground-state density functional theory (DFT) calculation. While Kohn-Sham DFT is used in about 40,000 scientific papers each year, the fundamental underpinnings are not widely appreciated. In this chapter, we analyze the inherent characteristics of DFT in their simplest form, using the asymmetric Hubbard dimer as an illustrative model. We begin by working through the core tenets of DFT, explaining what the exact ground-state density functional yields and does not yield. Given the relative simplicity of the system, almost all properties of the exact exchange-correlation functional are readily visualized and plotted. Key concepts include the Kohn-Sham scheme, the behavior of the XC potential as correlations become very strong, the derivative discontinuity and the difference between KS gaps and true charge gaps, and how to extract optical excitations using time-dependent DFT. By the end of this text and accompanying exercises, the reader will improve their ability to both explain and visualize the concepts of DFT, as well as better understand where others may go wrong.},
- archiveprefix = {arXiv},
+ pubstate = {preprint},
keywords = {DFT,physics,strongly correlated maeterials},
- file = {/home/johannes/Nextcloud/Zotero/Burke_Kozlowski_2021_Lies My Teacher Told Me About Density Functional Theory.pdf;/home/johannes/Zotero/storage/6EW6SVTP/2108.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Burke_Kozlowski_2021_Lies My Teacher Told Me About Density Functional Theory.pdf;/Users/wasmer/Zotero/storage/6EW6SVTP/2108.html}
}
@article{bystromCIDERExpressiveNonlocal2022,
@@ -1017,7 +1062,7 @@
urldate = {2022-05-11},
abstract = {Machine learning (ML) has recently gained attention as a means to develop more accurate exchange-correlation (XC) functionals for density functional theory, but functionals developed thus far need to be improved on several metrics, including accuracy, numerical stability, and transferability across chemical space. In this work, we introduce a set of nonlocal features of the density called the CIDER formalism, which we use to train a Gaussian process model for the exchange energy that obeys the critical uniform scaling rule for exchange. The resulting CIDER exchange functional is significantly more accurate than any semilocal functional tested here, and it has good transferability across main-group molecules. This work therefore serves as an initial step toward more accurate exchange functionals, and it also introduces useful techniques for developing robust, physics-informed XC models via ML.},
keywords = {DFT,ML,ML-DFT,ML-ESM},
- file = {/home/johannes/Nextcloud/Zotero/Bystrom_Kozinsky_2022_CIDER.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Bystrom_Kozinsky_2022_CIDER.pdf}
}
@article{calderonAFLOWStandardHighthroughput2015,
@@ -1035,7 +1080,7 @@
abstract = {The Automatic-Flow (AFLOW) standard for the high-throughput construction of materials science electronic structure databases is described. Electronic structure calculations of solid state materials depend on a large number of parameters which must be understood by researchers, and must be reported by originators to ensure reproducibility and enable collaborative database expansion. We therefore describe standard parameter values for k-point grid density, basis set plane wave kinetic energy cut-off, exchange–correlation functionals, pseudopotentials, DFT+U parameters, and convergence criteria used in AFLOW calculations.},
langid = {english},
keywords = {AFLOWLIB,High-throughput,Materials genomics,VASP},
- file = {/home/johannes/Nextcloud/Zotero/Calderon et al_2015_The AFLOW standard for high-throughput materials science calculations.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Calderon et al_2015_The AFLOW standard for high-throughput materials science calculations.pdf}
}
@article{cangiPotentialFunctionalsDensity2013,
@@ -1053,7 +1098,7 @@
urldate = {2022-07-08},
abstract = {Potential functional approximations are an intriguing alternative to density functional approximations. The potential functional that is dual to the Lieb density functional is defined and its properties are reported. The relationship between the Thomas-Fermi theory as a density functional and the theory as a potential functional is derived. The properties of several recent semiclassical potential functionals are explored, especially regarding their approach to the large particle number and classical continuum limits. The lack of ambiguity in the energy density of potential functional approximations is demonstrated. The density-density response function of the semiclassical approximation is calculated and shown to violate a key symmetry condition.},
keywords = {_tablet,density functional,density vs potential,DFT,potential},
- file = {/home/johannes/Nextcloud/Zotero/Cangi et al_2013_Potential functionals versus density functionals.pdf;/home/johannes/Zotero/storage/4U87YYPT/Cangi et al_2013_Potential functionals versus density functionals.pdf;/home/johannes/Zotero/storage/AJH43GTS/PhysRevA.88.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Cangi et al_2013_Potential functionals versus density functionals.pdf;/Users/wasmer/Zotero/storage/4U87YYPT/Cangi et al_2013_Potential functionals versus density functionals.pdf;/Users/wasmer/Zotero/storage/AJH43GTS/PhysRevA.88.html}
}
@article{caoArtificialIntelligenceHighthroughput2020,
@@ -1072,7 +1117,7 @@
urldate = {2021-05-21},
abstract = {Significant advances have been made in predicting new topological materials using high-throughput empirical descriptors or symmetry-based indicators. To date, these approaches have been applied to materials in existing databases, and are severely limited to systems with well-defined symmetries, leaving a much larger materials space unexplored. Using tetradymites as a prototypical class of examples, we uncover a two-dimensional descriptor by applying an artificial intelligence (AI)-based approach for fast and reliable identification of the topological characters of a drastically expanded range of materials, without prior determination of their specific symmetries and detailed band structures. By leveraging this descriptor that contains only the atomic number and electronegativity of the constituent species, we have readily scanned a huge number of alloys in the tetradymite family. Strikingly, nearly half of them are identified to be topological insulators, revealing a much larger territory of the topological materials world. The present work also attests to the increasingly important role of such AI-based approaches in modern materials discovery.},
keywords = {2D descriptor,classification,descriptors,GW approximation,HTC,materials discovery,materials screening,ML,SISSO,SVM,topological insulator},
- file = {/home/johannes/Nextcloud/Zotero/Cao et al_2020_Artificial intelligence for high-throughput discovery of topological insulators2.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Cao et al_2020_Artificial intelligence for high-throughput discovery of topological insulators2.pdf}
}
@article{caoTestsAccuracyScalability2020,
@@ -1093,7 +1138,7 @@
langid = {english},
pmcid = {PMC7746799},
keywords = {KKR},
- file = {/home/johannes/Nextcloud/Zotero/Cao et al_2020_Tests on the Accuracy and Scalability of the Full-Potential DFT Method Based on.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Cao et al_2020_Tests on the Accuracy and Scalability of the Full-Potential DFT Method Based on.pdf}
}
@unpublished{capelleBirdSeyeView2006,
@@ -1105,9 +1150,8 @@
url = {http://arxiv.org/abs/cond-mat/0211443},
urldate = {2021-08-31},
abstract = {This paper is the outgrowth of lectures the author gave at the Physics Institute and the Chemistry Institute of the University of Sao Paulo at Sao Carlos, Brazil, and at the VIII'th Summer School on Electronic Structure of the Brazilian Physical Society. It is an attempt to introduce density-functional theory (DFT) in a language accessible for students entering the field or researchers from other fields. It is not meant to be a scholarly review of DFT, but rather an informal guide to its conceptual basis and some recent developments and advances. The Hohenberg-Kohn theorem and the Kohn-Sham equations are discussed in some detail. Approximate density functionals, selected aspects of applications of DFT, and a variety of extensions of standard DFT are also discussed, albeit in less detail. Throughout it is attempted to provide a balanced treatment of aspects that are relevant for chemistry and aspects relevant for physics, but with a strong bias towards conceptual foundations. The paper is intended to be read before (or in parallel with) one of the many excellent more technical reviews available in the literature.},
- archiveprefix = {arXiv},
keywords = {_tablet,DFT,learn DFT,review},
- file = {/home/johannes/Nextcloud/Zotero/Capelle_2006_A bird's-eye view of density-functional theory.pdf;/home/johannes/Zotero/storage/8TLEU4M3/0211443.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Capelle_2006_A bird's-eye view of density-functional theory.pdf;/Users/wasmer/Zotero/storage/8TLEU4M3/0211443.html}
}
@article{carleoMachineLearningPhysical2019,
@@ -1120,7 +1164,7 @@
number = {4},
doi = {10.1103/RevModPhys.91.045002},
keywords = {Many-body theory,ML,review,science},
- file = {/home/johannes/Nextcloud/Zotero/Carleo_2019_Machine learning and the physical sciences.pdf;/home/johannes/Zotero/storage/9YE6JEBD/RevModPhys.91.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Carleo_2019_Machine learning and the physical sciences.pdf;/Users/wasmer/Zotero/storage/9YE6JEBD/RevModPhys.91.html}
}
@article{carleoSolvingQuantumManybody2017,
@@ -1136,7 +1180,7 @@
url = {https://www.science.org/doi/10.1126/science.aag2302},
urldate = {2022-03-29},
keywords = {ML,ML-QM,NN,rec-by-bluegel},
- file = {/home/johannes/Nextcloud/Zotero/Carleo_Troyer_2017_Solving the quantum many-body problem with artificial neural networks.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Carleo_Troyer_2017_Solving the quantum many-body problem with artificial neural networks.pdf}
}
@article{caroOptimizingManybodyAtomic2019,
@@ -1154,7 +1198,7 @@
urldate = {2021-05-13},
abstract = {We explore different ways to simplify the evaluation of the smooth overlap of atomic positions (SOAP) many-body atomic descriptor [Bartók et al., Phys. Rev. B 87, 184115 (2013).]. Our aim is to improve the computational efficiency of SOAP-based similarity kernel construction. While these improved atomic descriptors can be used for general characterization and interpolation of atomic properties, their main target application is accelerated evaluation of machine-learning-based interatomic potentials within the Gaussian approximation potential (GAP) framework [Bartók et al., Phys. Rev. Lett. 104, 136403 (2010)]. We achieve this objective by expressing the atomic densities in an approximate separable form, which decouples the radial and angular channels. We then express the elements of the SOAP descriptor (i.e., the expansion coefficients for the atomic densities) in analytical form given a particular choice of radial basis set. Finally, we derive recursion formulas for the expansion coefficients. This new SOAP-based descriptor allows for tenfold speedups compared to previous implementations, while improving the stability of the radial expansion for distant atomic neighbors, without degradation of the interpolation power of GAP models.},
keywords = {_tablet,descriptor dimred,descriptors,descriptors analysis,dimensionality reduction,GAP,ML,MLP,SOAP},
- file = {/home/johannes/Nextcloud/Zotero/Caro_2019_Optimizing many-body atomic descriptors for enhanced computational performance.pdf;/home/johannes/Zotero/storage/FDHHHJTR/PhysRevB.100.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Caro_2019_Optimizing many-body atomic descriptors for enhanced computational performance.pdf;/Users/wasmer/Zotero/storage/FDHHHJTR/PhysRevB.100.html}
}
@article{carvalhoRealspaceMappingTopological2018,
@@ -1172,14 +1216,13 @@
urldate = {2021-05-21},
abstract = {Topological invariants allow one to characterize Hamiltonians, predicting the existence of topologically protected in-gap modes. Those invariants can be computed by tracing the evolution of the occupied wave functions under twisted boundary conditions. However, those procedures do not allow one to calculate a topological invariant by evaluating the system locally, and thus require information about the wave functions in the whole system. Here we show that artificial neural networks can be trained to identify the topological order by evaluating a local projection of the density matrix. We demonstrate this for two different models, a one-dimensional topological superconductor and a two-dimensional quantum anomalous Hall state, both with spatially modulated parameters. Our neural network correctly identifies the different topological domains in real space, predicting the location of in-gap states. By combining a neural network with a calculation of the electronic states that uses the kernel polynomial method, we show that the local evaluation of the invariant can be carried out by evaluating a local quantity, in particular for systems without translational symmetry consisting of tens of thousands of atoms. Our results show that supervised learning is an efficient methodology to characterize the local topology of a system.},
keywords = {ANN,kernel methods,ML,superconductor,topological insulator,topological phase},
- file = {/home/johannes/Nextcloud/Zotero/Carvalho et al_2018_Real-space mapping of topological invariants using artificial neural networks.pdf;/home/johannes/Zotero/storage/ZNNT2KFN/PhysRevB.97.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Carvalho et al_2018_Real-space mapping of topological invariants using artificial neural networks.pdf;/Users/wasmer/Zotero/storage/ZNNT2KFN/PhysRevB.97.html}
}
@article{cavaIntroductionQuantumMaterials2021,
title = {Introduction: {{Quantum Materials}}},
shorttitle = {Introduction},
- author = {Cava, Robert and de Leon, Nathalie and Xie, Weiwei},
- options = {useprefix=true},
+ author = {Cava, Robert and family=Leon, given=Nathalie, prefix=de, useprefix=true and Xie, Weiwei},
date = {2021-03-10},
journaltitle = {Chemical Reviews},
shortjournal = {Chem. Rev.},
@@ -1191,7 +1234,7 @@
doi = {10.1021/acs.chemrev.0c01322},
url = {https://doi.org/10.1021/acs.chemrev.0c01322},
urldate = {2021-08-23},
- file = {/home/johannes/Nextcloud/Zotero/Cava et al_2021_Introduction.pdf;/home/johannes/Zotero/storage/EDPNL6EQ/acs.chemrev.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Cava et al_2021_Introduction.pdf;/Users/wasmer/Zotero/storage/EDPNL6EQ/acs.chemrev.html}
}
@article{ceriottiMachineLearningMeets2021,
@@ -1210,7 +1253,7 @@
urldate = {2021-05-13},
abstract = {Over recent years, the use of statistical learning techniques applied to chemical problems has gained substantial momentum. This is particularly apparent in the realm of physical chemistry, where the balance between empiricism and physics-based theory has traditionally been rather in favor of the latter. In this guest Editorial for the special topic issue on “Machine Learning Meets Chemical Physics,†a brief rationale is provided, followed by an overview of the topics covered. We conclude by making some general remarks.},
keywords = {ANN,Gaussian process,MD,ML,models,QC,review},
- file = {/home/johannes/Nextcloud/Zotero/Ceriotti et al_2021_Machine learning meets chemical physics.pdf;/home/johannes/Zotero/storage/6YB95LVA/5.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Ceriotti et al_2021_Machine learning meets chemical physics.pdf;/Users/wasmer/Zotero/storage/6YB95LVA/5.html}
}
@article{ceriottiPotentialsIntegratedMachine2022,
@@ -1226,8 +1269,8 @@
urldate = {2022-12-29},
abstract = {Over the past decade, interatomic potentials based on machine~learning (ML) techniques have become an indispensable tool in the atomic-scale modeling of materials. Trained on energies and forces obtained from electronic-structure calculations, they inherit their predictive accuracy, and extend greatly the length and time scales that are accessible to explicit atomistic simulations. Inexpensive predictions of the energetics of individual configurations have facilitated greatly the calculation of the thermodynamics of materials, including finite-temperature effects and disorder. More recently, ML models have been closing the gap with first-principles calculations in another area: the prediction of arbitrarily complicated functional properties, from vibrational and optical spectroscopies to electronic excitations. The implementation of integrated ML models that combine energetic and functional predictions with statistical and dynamical sampling of atomic-scale properties is bringing the promise of predictive, uncompromising simulations of existing and novel materials closer to its full realization.},
langid = {english},
- keywords = {equivariant,Gibbs free energy,integrated models,MLP,multiscale,prediction of DOS,prediction of polarizability,review,symmetry,tensorial target,thermodynamics},
- file = {/home/johannes/Nextcloud/Zotero/Ceriotti_2022_Beyond potentials.pdf}
+ keywords = {_tablet,equivariant,Gibbs free energy,integrated models,MLP,multiscale,prediction of DOS,prediction of polarizability,review,symmetry,tensorial target,thermodynamics},
+ file = {/Users/wasmer/Nextcloud/Zotero/Ceriotti_2022_Beyond potentials.pdf}
}
@article{ceriottiSimplifyingRepresentationComplex2011,
@@ -1249,7 +1292,7 @@
abstract = {A new scheme, sketch-map, for obtaining a low-dimensional representation of the region of phase space explored during an enhanced dynamics simulation is proposed. We show evidence, from an examination of the distribution of pairwise distances between frames, that some features of the free-energy surface are inherently high-dimensional. This makes dimensionality reduction problematic because the data does not satisfy the assumptions made in conventional manifold learning algorithms We therefore propose that when dimensionality reduction is performed on trajectory data one should think of the resultant embedding as a quickly sketched set of directions rather than a road map. In other words, the embedding tells one about the connectivity between states but does not provide the vectors that correspond to the slow degrees of freedom. This realization informs the development of sketch-map, which endeavors to reproduce the proximity information from the high-dimensionality description in a space of lower dimensionality even when a faithful embedding is not possible.},
langid = {english},
keywords = {data exploration,dimensionality reduction,library,MD,ML,sketchmap,unsupervised learning,visualization,with-code},
- file = {/home/johannes/Nextcloud/Zotero/Ceriotti et al_2011_Simplifying the representation of complex free-energy landscapes using.pdf;/home/johannes/Zotero/storage/GN7K44Z7/13023.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Ceriotti et al_2011_Simplifying the representation of complex free-energy landscapes using.pdf;/Users/wasmer/Zotero/storage/GN7K44Z7/13023.html}
}
@article{ceriottiUnsupervisedMachineLearning2019,
@@ -1268,7 +1311,23 @@
abstract = {Automated analyses of the outcome of a simulation have been an important part of atomistic modeling since the early days, addressing the need of linking the behavior of individual atoms and the collective properties that are usually the final quantity of interest. Methods such as clustering and dimensionality reduction have been used to provide a simplified, coarse-grained representation of the structure and dynamics of complex systems from proteins to nanoparticles. In recent years, the rise of machine learning has led to an even more widespread use of these algorithms in atomistic modeling and to consider different classification and inference techniques as part of a coherent toolbox of data-driven approaches. This perspective briefly reviews some of the unsupervised machine-learning methods—that are geared toward classification and coarse-graining of molecular simulations—seen in relation to the fundamental mathematical concepts that underlie all machine-learning techniques. It discusses the importance of using concise yet complete representations of atomic structures as the starting point of the analyses and highlights the risk of introducing preconceived biases when using machine learning to rationalize and understand structure-property relations. Supervised machine-learning techniques that explicitly attempt to predict the properties of a material given its structure are less susceptible to such biases. Current developments in the field suggest that using these two classes of approaches side-by-side and in a fully integrated mode, while keeping in mind the relations between the data analysis framework and the fundamental physical principles, will be key to realizing the full potential of machine learning to help understand the behavior of complex molecules and materials.},
langid = {english},
keywords = {clustering,data exploration,DBSCAN,dimensionality reduction,Gaussian mixture,ISOMAP,kernel methods,ML,multidimensional scaling,PCA,t-SNE,unsupervised learning},
- file = {/home/johannes/Nextcloud/Zotero/Ceriotti_2019_Unsupervised machine learning in atomistic simulations, between predictions and.pdf;/home/johannes/Zotero/storage/VYBXXJL4/1.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Ceriotti_2019_Unsupervised machine learning in atomistic simulations, between predictions and.pdf;/Users/wasmer/Zotero/storage/VYBXXJL4/1.html}
+}
+
+@online{cersonskyDatadrivenInterpretationStability2022,
+ title = {A Data-Driven Interpretation of the Stability of Molecular Crystals},
+ author = {Cersonsky, Rose K. and Pakhnova, Maria and Engel, Edgar A. and Ceriotti, Michele},
+ date = {2022-12-22},
+ number = {arXiv:2209.10709},
+ eprint = {arXiv:2209.10709},
+ eprinttype = {arxiv},
+ doi = {10.48550/arXiv.2209.10709},
+ url = {http://arxiv.org/abs/2209.10709},
+ urldate = {2023-01-23},
+ abstract = {Due to the subtle balance of intermolecular interactions that govern structure-property relations, predicting the stability of crystal structures formed from molecular building blocks is a highly non-trivial scientific problem. A particularly active and fruitful approach involves classifying the different combinations of interacting chemical moieties, as understanding the relative energetics of different interactions enables the design of molecular crystals and fine-tuning their stabilities. While this is usually performed based on the empirical observation of the most commonly encountered motifs in known crystal structures, we propose to apply a combination of supervised and unsupervised machine-learning techniques to automate the construction of an extensive library of molecular building blocks. We introduce a structural descriptor tailored to the prediction of the binding (lattice) energy and apply it to a curated dataset of organic crystals and exploit its atom-centered nature to obtain a data-driven assessment of the contribution of different chemical groups to the lattice energy of the crystal. We then interpret this library using a low-dimensional representation of the structure-energy landscape and discuss selected examples of the insights into crystal engineering that can be extracted from this analysis, providing a complete database to guide the design of molecular materials.},
+ pubstate = {preprint},
+ keywords = {_tablet,ACDC,crystal structure,dimensionality reduction,linear regression,organic crystals,PCovR,SOAP,structure prediction,structure search,unsupervised learning},
+ file = {/Users/wasmer/Nextcloud/Zotero/Cersonsky et al_2022_A data-driven interpretation of the stability of molecular crystals.pdf;/Users/wasmer/Zotero/storage/25LFMGQ5/2209.html}
}
@article{cersonskyImprovingSampleFeature2021,
@@ -1287,8 +1346,8 @@
urldate = {2022-08-10},
abstract = {Selecting the most relevant features and samples out of a large set of candidates is a task that occurs very often in the context of automated data analysis, where it improves the computational performance and often the transferability of a model. Here we focus on two popular subselection schemes applied to this end: CUR decomposition, derived from a low-rank approximation of the feature matrix, and farthest point sampling (FPS), which relies on the iterative identification of the most diverse samples and discriminating features. We modify these unsupervised approaches, incorporating a supervised component following the same spirit as the principal covariates (PCov) regression method. We show how this results in selections that perform better in supervised tasks, demonstrating with models of increasing complexity, from ridge regression to kernel ridge regression and finally feed-forward neural networks. We also present adjustments to minimise the impact of any subselection when performing unsupervised tasks. We demonstrate the significant improvements associated with PCov-CUR and PCov-FPS selections for applications to chemistry and materials science, typically reducing by a factor of two the number of features and samples required to achieve a given level of regression accuracy.},
langid = {english},
- keywords = {CUR decomposition,dimensionality reduction,feature selection,FPS,KRR,PCovR},
- file = {/home/johannes/Nextcloud/Zotero/Cersonsky et al_2021_Improving sample and feature selection with principal covariates regression.pdf;/home/johannes/Nextcloud/Zotero/false}
+ keywords = {_tablet,CUR decomposition,dimensionality reduction,feature selection,FPS,KPCovR,KRR,PCovR,sample selection},
+ file = {/home/johannes/Nextcloud/Zotero/false;/Users/wasmer/Nextcloud/Zotero/Cersonsky et al_2021_Improving sample and feature selection with principal covariates regression.pdf}
}
@article{chandrasekaranSolvingElectronicStructure2019,
@@ -1310,7 +1369,7 @@
langid = {english},
keywords = {_tablet,custom structural descriptors,descriptors,DFT,FCNN,grid-based descriptors,LDOS,ML,ML-DFT,ML-ESM,models,NN,prediction from structure,prediction of electron density,prediction of LDOS,RNN},
annotation = {Bandiera\_abtest: a Cc\_license\_type: cc\_by Cg\_type: Nature Research Journals Primary\_atype: Research Subject\_term: Computational methods;Electronic structure;Theory and computation Subject\_term\_id: computational-methods;electronic-structure;theory-and-computation},
- file = {/home/johannes/Nextcloud/Zotero/Chandrasekaran et al_2019_Solving the electronic structure problem with machine learning.pdf;/home/johannes/Nextcloud/Zotero/false;/home/johannes/Zotero/storage/TL92B668/s41524-019-0162-7.html}
+ file = {/home/johannes/Nextcloud/Zotero/false;/Users/wasmer/Nextcloud/Zotero/Chandrasekaran et al_2019_Solving the electronic structure problem with machine learning.pdf;/Users/wasmer/Zotero/storage/TL92B668/s41524-019-0162-7.html}
}
@article{changExperimentalObservationQuantum2013,
@@ -1325,7 +1384,7 @@
doi = {10.1126/science.1234414},
url = {https://www.science.org/doi/10.1126/science.1234414},
urldate = {2022-05-13},
- file = {/home/johannes/Nextcloud/Zotero/Chang et al_2013_Experimental Observation of the Quantum Anomalous Hall Effect in a Magnetic.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Chang et al_2013_Experimental Observation of the Quantum Anomalous Hall Effect in a Magnetic.pdf}
}
@unpublished{chardDLHubModelData2018,
@@ -1335,13 +1394,12 @@
date = {2018-11-27},
eprint = {1811.11213},
eprinttype = {arxiv},
- primaryclass = {cs, stat},
+ eprintclass = {cs, stat},
url = {http://arxiv.org/abs/1811.11213},
urldate = {2022-01-03},
abstract = {While the Machine Learning (ML) landscape is evolving rapidly, there has been a relative lag in the development of the "learning systems" needed to enable broad adoption. Furthermore, few such systems are designed to support the specialized requirements of scientific ML. Here we present the Data and Learning Hub for science (DLHub), a multi-tenant system that provides both model repository and serving capabilities with a focus on science applications. DLHub addresses two significant shortcomings in current systems. First, its selfservice model repository allows users to share, publish, verify, reproduce, and reuse models, and addresses concerns related to model reproducibility by packaging and distributing models and all constituent components. Second, it implements scalable and low-latency serving capabilities that can leverage parallel and distributed computing resources to democratize access to published models through a simple web interface. Unlike other model serving frameworks, DLHub can store and serve any Python 3-compatible model or processing function, plus multiple-function pipelines. We show that relative to other model serving systems including TensorFlow Serving, SageMaker, and Clipper, DLHub provides greater capabilities, comparable performance without memoization and batching, and significantly better performance when the latter two techniques can be employed. We also describe early uses of DLHub for scientific applications.},
- archiveprefix = {arXiv},
keywords = {Computer Science - Distributed; Parallel; and Cluster Computing,Computer Science - Machine Learning,Statistics - Machine Learning},
- file = {/home/johannes/Nextcloud/Zotero/Chard et al_2018_DLHub.pdf;/home/johannes/Zotero/storage/VT5H6PP6/1811.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Chard et al_2018_DLHub.pdf;/Users/wasmer/Zotero/storage/VT5H6PP6/1811.html}
}
@article{chenGraphNetworksUniversal2019,
@@ -1360,7 +1418,7 @@
urldate = {2022-01-02},
abstract = {Graph networks are a new machine learning (ML) paradigm that supports both relational reasoning and combinatorial generalization. Here, we develop universal MatErials Graph Network (MEGNet) models for accurate property prediction in both molecules and crystals. We demonstrate that the MEGNet models outperform prior ML models such as the SchNet in 11 out of 13 properties of the QM9 molecule data set. Similarly, we show that MEGNet models trained on ∼60 000 crystals in the Materials Project substantially outperform prior ML models in the prediction of the formation energies, band gaps, and elastic moduli of crystals, achieving better than density functional theory accuracy over a much larger data set. We present two new strategies to address data limitations common in materials science and chemistry. First, we demonstrate a physically intuitive approach to unify four separate molecular MEGNet models for the internal energy at 0 K and room temperature, enthalpy, and Gibbs free energy into a single free energy MEGNet model by incorporating the temperature, pressure, and entropy as global state inputs. Second, we show that the learned element embeddings in MEGNet models encode periodic chemical trends and can be transfer-learned from a property model trained on a larger data set (formation energies) to improve property models with smaller amounts of data (band gaps and elastic moduli).},
keywords = {_tablet,GNN,library,MEGNet,molecules,solids,vectorial learning target,with-code},
- file = {/home/johannes/Nextcloud/Zotero/Chen et al_2019_Graph Networks as a Universal Machine Learning Framework for Molecules and.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Chen et al_2019_Graph Networks as a Universal Machine Learning Framework for Molecules and.pdf}
}
@unpublished{chenUniversalGraphDeep2022,
@@ -1369,13 +1427,12 @@
date = {2022-02-04},
eprint = {2202.02450},
eprinttype = {arxiv},
- primaryclass = {cond-mat, physics:physics},
+ eprintclass = {cond-mat, physics:physics},
url = {http://arxiv.org/abs/2202.02450},
urldate = {2022-03-28},
abstract = {Interatomic potentials (IAPs), which describe the potential energy surface of a collection of atoms, are a fundamental input for atomistic simulations. However, existing IAPs are either fitted to narrow chemistries or too inaccurate for general applications. Here, we report a universal IAP for materials based on graph neural networks with three-body interactions (M3GNet). The M3GNet IAP was trained on the massive database of structural relaxations performed by the Materials Project over the past 10 years and has broad applications in structural relaxation, dynamic simulations and property prediction of materials across diverse chemical spaces. About 1.8 million potentially stable materials were identified from a screening of 31 million hypothetical crystal structures, demonstrating a machine learning-accelerated pathway to the discovery of synthesizable materials with exceptional properties.},
- archiveprefix = {arXiv},
keywords = {_tablet,condensed matter,GNN,library,M3GNet,materials,materials database,materials project,matterverse,MEGNet,ML,MLP,molecules,periodic table,solids,tensorial target,with-code},
- file = {/home/johannes/Nextcloud/Zotero/Chen_Ong_2022_A Universal Graph Deep Learning Interatomic Potential for the Periodic Table.pdf;/home/johannes/Zotero/storage/H4FKVKUF/2202.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Chen_Ong_2022_A Universal Graph Deep Learning Interatomic Potential for the Periodic Table.pdf;/Users/wasmer/Zotero/storage/H4FKVKUF/2202.html}
}
@article{choudharyAtomisticLineGraph2021,
@@ -1396,7 +1453,7 @@
issue = {1},
langid = {english},
keywords = {ALIGNN,GNN,ML,MPNN,original publication},
- file = {/home/johannes/Nextcloud/Zotero/Choudhary_DeCost_2021_Atomistic Line Graph Neural Network for improved materials property predictions.pdf;/home/johannes/Zotero/storage/F8XSYTPV/s41524-021-00650-1.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Choudhary_DeCost_2021_Atomistic Line Graph Neural Network for improved materials property predictions.pdf;/Users/wasmer/Zotero/storage/F8XSYTPV/s41524-021-00650-1.html}
}
@article{choudharyMachineLearningForcefieldinspired2018,
@@ -1415,7 +1472,7 @@
urldate = {2021-06-26},
abstract = {We present a complete set of chemo-structural descriptors to significantly extend the applicability of machine learning (ML) in material screening and mapping the energy landscape for multicomponent systems. These descriptors allow differentiating between structural prototypes, which is not possible using the commonly used chemical-only descriptors. Specifically, we demonstrate that the combination of pairwise radial, nearest-neighbor, bond-angle, dihedral-angle, and core-charge distributions plays an important role in predicting formation energies, band gaps, static refractive indices, magnetic properties, and modulus of elasticity for three-dimensional materials as well as exfoliation energies of two-dimensional (2D)-layered materials. The training data consist of 24 549 bulk and 616 monolayer materials taken from the JARVIS-DFT database. We obtained very accurate ML models using a gradient-boosting algorithm. Then we use the trained models to discover exfoliable 2D-layered materials satisfying specific property requirements. Additionally, we integrate our formation-energy ML model with a genetic algorithm for structure search to verify if the ML model reproduces the density-functional-theory convex hull. This verification establishes a more stringent evaluation metric for the ML model than what is commonly used in data sciences. Our learned model is publicly available on the JARVIS-ML website (https://www.ctcms.nist.gov/jarvisml), property predictions of generalized materials.},
keywords = {A3MDNet,CFID,classification of magnetic/nonmagnetic,classification of metal/insulator,descriptors,DFT,GBDT,JARVIS,ML,models},
- file = {/home/johannes/Nextcloud/Zotero/Choudhary et al_2018_Machine learning with force-field-inspired descriptors for materials Author's Manuscript.pdf;/home/johannes/Nextcloud/Zotero/Choudhary et al_2018_Machine learning with force-field-inspired descriptors for materials Suppl JARVIS-ML.pdf;/home/johannes/Nextcloud/Zotero/Choudhary et al_2018_Machine learning with force-field-inspired descriptors for materials.pdf;/home/johannes/Zotero/storage/88LWP9IL/Choudhary et al_2018_Machine learning with force-field-inspired descriptors for materials Suppl Feature Importance.xlsx;/home/johannes/Zotero/storage/8U5VA8X6/Choudhary et al_2018_Machine learning with force-field-inspired descriptors for materials Suppl Feature Importance.xlsx;/home/johannes/Zotero/storage/NYHDKNR3/PhysRevMaterials.2.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Choudhary et al_2018_Machine learning with force-field-inspired descriptors for materials Author's Manuscript.pdf;/Users/wasmer/Nextcloud/Zotero/Choudhary et al_2018_Machine learning with force-field-inspired descriptors for materials Suppl JARVIS-ML.pdf;/Users/wasmer/Nextcloud/Zotero/Choudhary et al_2018_Machine learning with force-field-inspired descriptors for materials.pdf;/Users/wasmer/Zotero/storage/88LWP9IL/Choudhary et al_2018_Machine learning with force-field-inspired descriptors for materials Suppl Feature Importance.xlsx;/Users/wasmer/Zotero/storage/8U5VA8X6/Choudhary et al_2018_Machine learning with force-field-inspired descriptors for materials Suppl Feature Importance.xlsx;/Users/wasmer/Zotero/storage/NYHDKNR3/PhysRevMaterials.2.html}
}
@unpublished{chouhanImprovingScalabilityReliability2021,
@@ -1424,13 +1481,32 @@
date = {2021-03-16},
eprint = {2103.08546},
eprinttype = {arxiv},
- primaryclass = {cs},
+ eprintclass = {cs},
url = {http://arxiv.org/abs/2103.08546},
urldate = {2021-10-20},
abstract = {Checkpoint/restart (C/R) provides fault-tolerant computing capability, enables long running applications, and provides scheduling flexibility for computing centers to support diverse workloads with different priority. It is therefore vital to get transparent C/R capability working at NERSC. MANA, by Garg et. al., is a transparent checkpointing tool that has been selected due to its MPI-agnostic and network-agnostic approach. However, originally written as a proof-of-concept code, MANA was not ready to use with NERSC's diverse production workloads, which are dominated by MPI and hybrid MPI+OpenMP applications. In this talk, we present ongoing work at NERSC to enable MANA for NERSC's production workloads, including fixing bugs that were exposed by the top applications at NERSC, adding new features to address system changes, evaluating C/R overhead at scale, etc. The lessons learned from making MANA production-ready for HPC applications will be useful for C/R tool developers, supercomputing centers and HPC end-users alike.},
- archiveprefix = {arXiv},
keywords = {DFT,for introductions,master-thesis,supercomputing},
- file = {/home/johannes/Nextcloud/Zotero/Chouhan et al_2021_Improving scalability and reliability of MPI-agnostic transparent checkpointing.pdf;/home/johannes/Zotero/storage/BTL4HHE6/2103.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Chouhan et al_2021_Improving scalability and reliability of MPI-agnostic transparent checkpointing.pdf;/Users/wasmer/Zotero/storage/BTL4HHE6/2103.html}
+}
+
+@article{christensen2022RoadmapNeuromorphic2022,
+ title = {2022 Roadmap on Neuromorphic Computing and Engineering},
+ author = {Christensen, Dennis V. and Dittmann, Regina and Linares-Barranco, Bernabe and Sebastian, Abu and Gallo, Manuel Le and Redaelli, Andrea and Slesazeck, Stefan and Mikolajick, Thomas and Spiga, Sabina and Menzel, Stephan and Valov, Ilia and Milano, Gianluca and Ricciardi, Carlo and Liang, Shi-Jun and Miao, Feng and Lanza, Mario and Quill, Tyler J. and Keene, Scott T. and Salleo, Alberto and Grollier, Julie and Marković, Danijela and Mizrahi, Alice and Yao, Peng and Yang, J. Joshua and Indiveri, Giacomo and Strachan, John Paul and Datta, Suman and Vianello, Elisa and Valentian, Alexandre and Feldmann, Johannes and Li, Xuan and Pernice, Wolfram H. P. and Bhaskaran, Harish and Furber, Steve and Neftci, Emre and Scherr, Franz and Maass, Wolfgang and Ramaswamy, Srikanth and Tapson, Jonathan and Panda, Priyadarshini and Kim, Youngeun and Tanaka, Gouhei and Thorpe, Simon and Bartolozzi, Chiara and Cleland, Thomas A. and Posch, Christoph and Liu, ShihChii and Panuccio, Gabriella and Mahmud, Mufti and Mazumder, Arnab Neelim and Hosseini, Morteza and Mohsenin, Tinoosh and Donati, Elisa and Tolu, Silvia and Galeazzi, Roberto and Christensen, Martin Ejsing and Holm, Sune and Ielmini, Daniele and Pryds, N.},
+ date = {2022-05},
+ journaltitle = {Neuromorphic Computing and Engineering},
+ shortjournal = {Neuromorph. Comput. Eng.},
+ volume = {2},
+ number = {2},
+ pages = {022501},
+ publisher = {{IOP Publishing}},
+ issn = {2634-4386},
+ doi = {10.1088/2634-4386/ac4a83},
+ url = {https://dx.doi.org/10.1088/2634-4386/ac4a83},
+ urldate = {2023-03-06},
+ abstract = {Modern computation based on von Neumann architecture is now a mature cutting-edge science. In the von Neumann architecture, processing and memory units are implemented as separate blocks interchanging data intensively and continuously. This data transfer is responsible for a large part of the power consumption. The next generation computer technology is expected to solve problems at the exascale with 1018 calculations each second. Even though these future computers will be incredibly powerful, if they are based on von Neumann type architectures, they will consume between 20 and 30 megawatts of power and will not have intrinsic physically built-in capabilities to learn or deal with complex data as our brain does. These needs can be addressed by neuromorphic computing systems which are inspired by the biological concepts of the human brain. This new generation of computers has the potential to be used for the storage and processing of large amounts of digital information with much lower power consumption than conventional processors. Among their potential future applications, an important niche is moving the control from data centers to edge devices. The aim of this roadmap is to present a snapshot of the present state of neuromorphic technology and provide an opinion on the challenges and opportunities that the future holds in the major areas of neuromorphic technology, namely materials, devices, neuromorphic circuits, neuromorphic algorithms, applications, and ethics. The roadmap is a collection of perspectives where leading researchers in the neuromorphic community provide their own view about the current state and the future challenges for each research area. We hope that this roadmap will be a useful resource by providing a concise yet comprehensive introduction to readers outside this field, for those who are just entering the field, as well as providing future perspectives for those who are well established in the neuromorphic computing community.},
+ langid = {english},
+ keywords = {/unread,computing,Deep learning,for introductions,Neuromorphic,NN,review,Spintronics,unconventional computing},
+ file = {/Users/wasmer/Nextcloud/Zotero/Christensen et al_2022_2022 roadmap on neuromorphic computing and engineering.pdf}
}
@article{clementBenchmarkAFLOWData2020,
@@ -1448,7 +1524,7 @@
urldate = {2021-10-15},
abstract = {Materials informatics is increasingly finding ways to exploit machine learning algorithms. Techniques such as decision trees, ensemble methods, support vector machines, and a variety of neural network architectures are used to predict likely material characteristics and property values. Supplemented with laboratory synthesis, applications of machine learning to compound discovery and characterization represent one of the most promising research directions in materials informatics. A shortcoming of this trend, in its current form, is a lack of standardized materials data sets on which to train, validate, and test model effectiveness. Applied machine learning research depends on benchmark data to make sense of its results. Fixed, predetermined data sets allow for rigorous model assessment and comparison. Machine learning publications that do not refer to benchmarks are often hard to contextualize and reproduce. In this data descriptor article, we present a collection of data sets of different material properties taken from the AFLOW database. We describe them, the procedures that generated them, and their use as potential benchmarks. We provide a compressed ZIP file containing the data sets and a GitHub repository of associated Python code. Finally, we discuss opportunities for future work incorporating the data sets and creating similar benchmark collections.},
langid = {english},
- file = {/home/johannes/Nextcloud/Zotero/Clement et al_2020_Benchmark AFLOW Data Sets for Machine Learning.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Clement et al_2020_Benchmark AFLOW Data Sets for Machine Learning.pdf}
}
@unpublished{cobelliInversionChemicalEnvironment2022,
@@ -1457,13 +1533,12 @@
date = {2022-01-27},
eprint = {2201.11591},
eprinttype = {arxiv},
- primaryclass = {cond-mat},
+ eprintclass = {cond-mat},
url = {http://arxiv.org/abs/2201.11591},
urldate = {2022-03-23},
abstract = {Machine-learning generative methods for material design are constructed by representing a given chemical structure, either a solid or a molecule, over appropriate atomic features, generally called structural descriptors. These must be fully descriptive of the system, must facilitate the training process and must be invertible, so that one can extract the atomic configurations corresponding to the output of the model. In general, this last requirement is not automatically satisfied by the most efficient structural descriptors, namely the representation is not directly invertible. Such drawback severely limits our freedom of choice in selecting the most appropriate descriptors for the problem, and thus our flexibility to construct generative models. In this work, we present a general optimization method capable of inverting any local many-body descriptor of the chemical environment, back to a cartesian representation. The algorithm is then implemented together with the bispectrum representation of the local structure and demonstrated for a number of molecules. The scheme presented here, thus, represents a general approach to the inversion of structural descriptors, enabling the construction of efficient structural generative models.},
- archiveprefix = {arXiv},
- keywords = {descriptors,generative models,inversion,ML},
- file = {/home/johannes/Nextcloud/Zotero/Cobelli et al_2022_Inversion of the chemical environment representations.pdf;/home/johannes/Zotero/storage/A6MH6ZIG/2201.html}
+ keywords = {_tablet,descriptors,generative models,inversion,ML},
+ file = {/Users/wasmer/Nextcloud/Zotero/Cobelli et al_2022_Inversion of the chemical environment representations.pdf;/Users/wasmer/Zotero/storage/A6MH6ZIG/2201.html}
}
@article{collinsHumanGenomeProject2003,
@@ -1479,7 +1554,7 @@
doi = {10.1126/science.1084564},
url = {https://www.science.org/doi/full/10.1126/science.1084564},
urldate = {2021-10-15},
- file = {/home/johannes/Nextcloud/Zotero/Collins et al_2003_The Human Genome Project.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Collins et al_2003_The Human Genome Project.pdf}
}
@article{cuevas-zuviriaMachineLearningAnalytical2021,
@@ -1498,7 +1573,7 @@
urldate = {2022-07-10},
abstract = {Machine learning milestones in computational chemistry are overshadowed by their unaccountability and the overwhelming zoo of tools for each specific task. A promising path to tackle these problems is using machine learning to reproduce physical magnitudes as a basis to derive many other properties. By using a model of the electron density consisting of an analytical expansion on a linear set of isotropic and anisotropic functions, we implemented in this work a message-passing neural network able to reproduce electron density in molecules with just a 2.5\% absolute error in complex cases. We also adapted our methodology to describe electron density in large biomolecules (proteins) and to obtain atomic charges, interaction energies, and DFT energies. We show that electron density learning is a new promising avenue with a variety of forthcoming applications.},
keywords = {analytical model,GCN,GNN,ML,molecules,MPNN,prediction of electron density},
- file = {/home/johannes/Nextcloud/Zotero/Cuevas-ZuvirÃa_Pacios_2021_Machine Learning of Analytical Electron Density in Large Molecules Through.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Cuevas-ZuvirÃa_Pacios_2021_Machine Learning of Analytical Electron Density in Large Molecules Through.pdf}
}
@article{curtaroloAFLOWAutomaticFramework2012,
@@ -1517,7 +1592,7 @@
abstract = {Recent advances in computational materials science present novel opportunities for structure discovery and optimization, including uncovering of unsuspected compounds and metastable structures, electronic structure, surface, and nano-particle properties. The practical realization of these opportunities requires systematic generation and classification of the relevant computational data by high-throughput methods. In this paper we present Aflow (Automatic Flow), a software framework for high-throughput calculation of crystal structure properties of alloys, intermetallics and inorganic compounds. The Aflow software is available for the scientific community on the website of the materials research consortium, aflowlib.org. Its geometric and electronic structure analysis and manipulation tools are additionally available for online operation at the same website. The combination of automatic methods and user online interfaces provide a powerful tool for efficient quantum computational materials discovery and characterization.},
langid = {english},
keywords = {Ab initio,AFLOW,Combinatorial materials science,High-throughput},
- file = {/home/johannes/Nextcloud/Zotero/Curtarolo et al_2012_AFLOW.pdf;/home/johannes/Zotero/storage/3ZKE8YHP/S0927025612000717.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Curtarolo et al_2012_AFLOW.pdf;/Users/wasmer/Zotero/storage/3ZKE8YHP/S0927025612000717.html}
}
@article{curtaroloAFLOWLIBORGDistributed2012,
@@ -1536,7 +1611,7 @@
abstract = {Empirical databases of crystal structures and thermodynamic properties are fundamental tools for materials research. Recent rapid proliferation of computational data on materials properties presents the possibility to complement and extend the databases where the experimental data is lacking or difficult to obtain. Enhanced repositories that integrate both computational and empirical approaches open novel opportunities for structure discovery and optimization, including uncovering of unsuspected compounds, metastable structures and correlations between various characteristics. The practical realization of these opportunities depends on a systematic compilation and classification of the generated data in addition to an accessible interface for the materials science community. In this paper we present an extensive repository, aflowlib.org, comprising phase-diagrams, electronic structure and magnetic properties, generated by the high-throughput framework AFLOW. This continuously updated compilation currently contains over 150,000 thermodynamic entries for alloys, covering the entire composition range of more than 650 binary systems, 13,000 electronic structure analyses of inorganic compounds, and 50,000 entries for novel potential magnetic and spintronics systems. The repository is available for the scientific community on the website of the materials research consortium, aflowlib.org.},
langid = {english},
keywords = {Ab initio,AFLOW,Combinatorial materials science,High-throughput,Materials genome initiative},
- file = {/home/johannes/Zotero/storage/5MTYTHXV/S0927025612000687.html}
+ file = {/Users/wasmer/Zotero/storage/5MTYTHXV/S0927025612000687.html}
}
@unpublished{darbyCompressingLocalAtomic2021,
@@ -1545,13 +1620,12 @@
date = {2021-12-24},
eprint = {2112.13055},
eprinttype = {arxiv},
- primaryclass = {cond-mat},
+ eprintclass = {cond-mat},
url = {http://arxiv.org/abs/2112.13055},
urldate = {2022-01-03},
abstract = {Many atomic descriptors are currently limited by their unfavourable scaling with the number of chemical elements \$S\$ e.g. the length of body-ordered descriptors, such as the Smooth Overlap of Atomic Positions (SOAP) power spectrum (3-body) and the Atomic Cluster Expansion (ACE) (multiple body-orders), scales as \$(NS)\^\textbackslash nu\$ where \$\textbackslash nu+1\$ is the body-order and \$N\$ is the number of radial basis functions used in the density expansion. We introduce two distinct approaches which can be used to overcome this scaling for the SOAP power spectrum. Firstly, we show that the power spectrum is amenable to lossless compression with respect to both \$S\$ and \$N\$, so that the descriptor length can be reduced from \$\textbackslash mathcal\{O\}(N\^2S\^2)\$ to \$\textbackslash mathcal\{O\}\textbackslash left(NS\textbackslash right)\$. Secondly, we introduce a generalized SOAP kernel, where compression is achieved through the use of the total, element agnostic density, in combination with radial projection. The ideas used in the generalized kernel are equally applicably to any other body-ordered descriptors and we demonstrate this for the Atom Centered Symmetry Functions (ACSF). Finally, both compression approaches are shown to offer comparable performance to the original descriptor across a variety of numerical tests.},
- archiveprefix = {arXiv},
keywords = {_tablet,ACE,ACSF,chemical species scaling problem,descriptor dimred,descriptors,descriptors analysis,dimensionality reduction,library,ML,SOAP,with-code},
- file = {/home/johannes/Nextcloud/Zotero/Darby et al_2021_Compressing local atomic neighbourhood descriptors.pdf;/home/johannes/Zotero/storage/GXXQQPAA/2112.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Darby et al_2021_Compressing local atomic neighbourhood descriptors.pdf;/Users/wasmer/Zotero/storage/GXXQQPAA/2112.html}
}
@article{darbyCompressingLocalAtomic2022,
@@ -1571,26 +1645,24 @@
abstract = {Many atomic descriptors are currently limited by their unfavourable scaling with the number of chemical elements S e.g. the length of body-ordered descriptors, such as the SOAP power spectrum (3-body) and the (ACE) (multiple body-orders), scales as (NS)ν where ν\,+\,1 is the body-order and N is the number of radial basis functions used in the density expansion. We introduce two distinct approaches which can be used to overcome this scaling for the SOAP power spectrum. Firstly, we show that the power spectrum is amenable to lossless compression with respect to both S and N, so that the descriptor length can be reduced from \$\$\{\{\{\textbackslash mathcal\{O\}\}\}\}(\{N\}\^\{2\}\{S\}\^\{2\})\$\$to \$\$\{\{\{\textbackslash mathcal\{O\}\}\}\}\textbackslash left(NS\textbackslash right)\$\$. Secondly, we introduce a generalised SOAP kernel, where compression is achieved through the use of the total, element agnostic density, in combination with radial projection. The ideas used in the generalised kernel are equally applicably to any other body-ordered descriptors and we demonstrate this for the (ACSF).},
issue = {1},
langid = {english},
- keywords = {ACE,ACSF,chemical species scaling problem,descriptor dimred,descriptors,descriptors analysis,dimensionality reduction,library,ML,SOAP,with-code},
- file = {/home/johannes/Nextcloud/Zotero/Darby et al_2022_Compressing local atomic neighbourhood descriptors.pdf;/home/johannes/Zotero/storage/WR6IJ7MC/s41524-022-00847-y.html}
+ keywords = {_tablet,ACE,ACSF,chemical species scaling problem,descriptor dimred,descriptors,descriptors analysis,dimensionality reduction,library,ML,SOAP,with-code},
+ file = {/Users/wasmer/Nextcloud/Zotero/Darby et al_2022_Compressing local atomic neighbourhood descriptors.pdf;/Users/wasmer/Zotero/storage/WR6IJ7MC/s41524-022-00847-y.html}
}
-@misc{darbyTensorreducedAtomicDensity2022,
+@online{darbyTensorreducedAtomicDensity2022,
title = {Tensor-Reduced Atomic Density Representations},
author = {Darby, James P. and Kovács, Dávid P. and Batatia, Ilyes and Caro, Miguel A. and Hart, Gus L. W. and Ortner, Christoph and Csányi, Gábor},
date = {2022-10-01},
number = {arXiv:2210.01705},
- eprint = {2210.01705},
+ eprint = {arXiv:2210.01705},
eprinttype = {arxiv},
- primaryclass = {cond-mat, physics:physics},
- publisher = {{arXiv}},
doi = {10.48550/arXiv.2210.01705},
url = {http://arxiv.org/abs/2210.01705},
urldate = {2022-10-05},
abstract = {Density based representations of atomic environments that are invariant under Euclidean symmetries have become a widely used tool in the machine learning of interatomic potentials, broader data-driven atomistic modelling and the visualisation and analysis of materials datasets.The standard mechanism used to incorporate chemical element information is to create separate densities for each element and form tensor products between them. This leads to a steep scaling in the size of the representation as the number of elements increases. Graph neural networks, which do not explicitly use density representations, escape this scaling by mapping the chemical element information into a fixed dimensional space in a learnable way. We recast this approach as tensor factorisation by exploiting the tensor structure of standard neighbour density based descriptors. In doing so, we form compact tensor-reduced representations whose size does not depend on the number of chemical elements, but remain systematically convergeable and are therefore applicable to a wide range of data analysis and regression tasks.},
- archiveprefix = {arXiv},
- keywords = {ACE,chemical species scaling problem,descriptor dimred,descriptors,dimensionality reduction,MACE,ML,Multi-ACE},
- file = {/home/johannes/Nextcloud/Zotero/Darby et al_2022_Tensor-reduced atomic density representations.pdf;/home/johannes/Zotero/storage/6XMXCLL4/2210.html}
+ pubstate = {preprint},
+ keywords = {_tablet,ACE,chemical species scaling problem,descriptor dimred,descriptors,dimensionality reduction,MACE,ML,Multi-ACE},
+ file = {/Users/wasmer/Nextcloud/Zotero/Darby et al_2022_Tensor-reduced atomic density representations.pdf;/Users/wasmer/Zotero/storage/6XMXCLL4/2210.html}
}
@article{dasCrysXPPExplainableProperty2022,
@@ -1612,7 +1684,7 @@
issue = {1},
langid = {english},
keywords = {autoencoder,CGCNN,CrysXPP,dimensionality reduction,feature selection,GCN,GNN,library,solids,with-code},
- file = {/home/johannes/Nextcloud/Zotero/Das et al_2022_CrysXPP.pdf;/home/johannes/Zotero/storage/SY9ITHM4/s41524-022-00716-8.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Das et al_2022_CrysXPP.pdf;/Users/wasmer/Zotero/storage/SY9ITHM4/s41524-022-00716-8.html}
}
@inproceedings{davidsonProvenanceScientificWorkflows2008,
@@ -1631,7 +1703,7 @@
abstract = {Provenance in the context of workflows, both for the data they derive and for their specification, is an essential component to allow for result reproducibility, sharing, and knowledge re-use in the scientific community. Several workshops have been held on the topic, and it has been the focus of many research projects and prototype systems. This tutorial provides an overview of research issues in provenance for scientific workflows, with a focus on recent literature and technology in this area. It is aimed at a general database research audience and at people who work with scientific data and workflows. We will (1) provide a general overview of scientific workflows, (2) describe research on provenance for scientific workflows and show in detail how provenance is supported in existing systems; (3) discuss emerging applications that are enabled by provenance; and (4) outline open problems and new directions for database-related research.},
isbn = {978-1-60558-102-6},
keywords = {provenance,scientific workflows},
- file = {/home/johannes/Nextcloud/Zotero/Davidson_Freire_2008_Provenance and scientific workflows.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Davidson_Freire_2008_Provenance and scientific workflows.pdf}
}
@article{deComparingMoleculesSolids2016,
@@ -1651,13 +1723,12 @@
abstract = {Evaluating the (dis)similarity of crystalline, disordered and molecular compounds is a critical step in the development of algorithms to navigate automatically the configuration space of complex materials. For instance, a structural similarity metric is crucial for classifying structures, searching chemical space for better compounds and materials, and driving the next generation of machine-learning techniques for predicting the stability and properties of molecules and materials. In the last few years several strategies have been designed to compare atomic coordination environments. In particular, the smooth overlap of atomic positions (SOAPs) has emerged as an elegant framework to obtain translation, rotation and permutation-invariant descriptors of groups of atoms, underlying the development of various classes of machine-learned inter-atomic potentials. Here we discuss how one can combine such local descriptors using a regularized entropy match (REMatch) approach to describe the similarity of both whole molecular and bulk periodic structures, introducing powerful metrics that enable the navigation of alchemical and structural complexities within a unified framework. Furthermore, using this kernel and a ridge regression method we can predict atomization energies for a database of small organic molecules with a mean absolute error below 1 kcal mol−1, reaching an important milestone in the application of machine-learning techniques for the evaluation of molecular properties.},
langid = {english},
keywords = {classification,descriptors,kernel methods,ML,rec-by-tim-wuerger,REMatch,similarity analysis,SOAP},
- file = {/home/johannes/Nextcloud/Zotero/De et al_2016_Comparing molecules and solids across structural and alchemical space.pdf;/home/johannes/Zotero/storage/SA8QCH28/C6CP00415F.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/De et al_2016_Comparing molecules and solids across structural and alchemical space.pdf;/Users/wasmer/Zotero/storage/SA8QCH28/C6CP00415F.html}
}
@article{degraveMagneticControlTokamak2022,
title = {Magnetic Control of Tokamak Plasmas through Deep Reinforcement Learning},
- author = {Degrave, Jonas and Felici, Federico and Buchli, Jonas and Neunert, Michael and Tracey, Brendan and Carpanese, Francesco and Ewalds, Timo and Hafner, Roland and Abdolmaleki, Abbas and de las Casas, Diego and Donner, Craig and Fritz, Leslie and Galperti, Cristian and Huber, Andrea and Keeling, James and Tsimpoukelli, Maria and Kay, Jackie and Merle, Antoine and Moret, Jean-Marc and Noury, Seb and Pesamosca, Federico and Pfau, David and Sauter, Olivier and Sommariva, Cristian and Coda, Stefano and Duval, Basil and Fasoli, Ambrogio and Kohli, Pushmeet and Kavukcuoglu, Koray and Hassabis, Demis and Riedmiller, Martin},
- options = {useprefix=true},
+ author = {Degrave, Jonas and Felici, Federico and Buchli, Jonas and Neunert, Michael and Tracey, Brendan and Carpanese, Francesco and Ewalds, Timo and Hafner, Roland and Abdolmaleki, Abbas and family=Casas, given=Diego, prefix=de las, useprefix=true and Donner, Craig and Fritz, Leslie and Galperti, Cristian and Huber, Andrea and Keeling, James and Tsimpoukelli, Maria and Kay, Jackie and Merle, Antoine and Moret, Jean-Marc and Noury, Seb and Pesamosca, Federico and Pfau, David and Sauter, Olivier and Sommariva, Cristian and Coda, Stefano and Duval, Basil and Fasoli, Ambrogio and Kohli, Pushmeet and Kavukcuoglu, Koray and Hassabis, Demis and Riedmiller, Martin},
date = {2022-02},
journaltitle = {Nature},
volume = {602},
@@ -1672,7 +1743,7 @@
issue = {7897},
langid = {english},
keywords = {Computer science,Magnetically confined plasmas,ML,Nuclear fusion and fission,rec-by-bluegel},
- file = {/home/johannes/Nextcloud/Zotero/Degrave et al_2022_Magnetic control of tokamak plasmas through deep reinforcement learning.pdf;/home/johannes/Zotero/storage/U6PRS6KM/s41586-021-04301-9.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Degrave et al_2022_Magnetic control of tokamak plasmas through deep reinforcement learning.pdf;/Users/wasmer/Zotero/storage/U6PRS6KM/s41586-021-04301-9.html}
}
@article{dengImageNetLargescaleHierarchical2009,
@@ -1700,7 +1771,25 @@
urldate = {2021-05-21},
abstract = {Artificial neural networks and machine learning have now reached a new era after several decades of improvement where applications are to explode in many fields of science, industry, and technology. Here, we use artificial neural networks to study an intriguing phenomenon in quantum physics—the topological phases of matter. We find that certain topological states, either symmetry-protected or with intrinsic topological order, can be represented with classical artificial neural networks. This is demonstrated by using three concrete spin systems, the one-dimensional (1D) symmetry-protected topological cluster state and the 2D and 3D toric code states with intrinsic topological orders. For all three cases, we show rigorously that the topological ground states can be represented by short-range neural networks in an exact and efficient fashion—the required number of hidden neurons is as small as the number of physical spins and the number of parameters scales only linearly with the system size. For the 2D toric-code model, we find that the proposed short-range neural networks can describe the excited states with Abelian anyons and their nontrivial mutual statistics as well. In addition, by using reinforcement learning we show that neural networks are capable of finding the topological ground states of nonintegrable Hamiltonians with strong interactions and studying their topological phase transitions. Our results demonstrate explicitly the exceptional power of neural networks in describing topological quantum states, and at the same time provide valuable guidance to machine learning of topological phases in generic lattice models.},
keywords = {ANN,ML,reinforcement-learning,topological phase,topological phase transition},
- file = {/home/johannes/Nextcloud/Zotero/Deng et al_2017_Machine learning topological states.pdf;/home/johannes/Zotero/storage/QSGREI8E/PhysRevB.96.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Deng et al_2017_Machine learning topological states.pdf;/Users/wasmer/Zotero/storage/QSGREI8E/PhysRevB.96.html}
+}
+
+@article{dengMeltingMathrmMgSiMathrmO2023,
+ title = {Melting of \$\textbackslash mathrm\{\vphantom\}{{MgSi}}\vphantom\{\}\{\textbackslash mathrm\{\vphantom{\}\}}{{O}}\vphantom\{\}\vphantom\{\}\_\{3\}\$ Determined by Machine Learning Potentials},
+ author = {Deng, Jie and Niu, Haiyang and Hu, Junwei and Chen, Mingyi and Stixrude, Lars},
+ date = {2023-02-13},
+ journaltitle = {Physical Review B},
+ shortjournal = {Phys. Rev. B},
+ volume = {107},
+ number = {6},
+ pages = {064103},
+ publisher = {{American Physical Society}},
+ doi = {10.1103/PhysRevB.107.064103},
+ url = {https://link.aps.org/doi/10.1103/PhysRevB.107.064103},
+ urldate = {2023-03-01},
+ abstract = {Melting in the deep rocky portions of planets is important for understanding the thermal evolution of these bodies and the possible generation of magnetic fields in their underlying metallic cores. But the melting temperature of silicates is poorly constrained at the pressures expected in super-Earth exoplanets, the most abundant type of planets in the galaxy. Here, we propose an iterative learning scheme that combines enhanced sampling, feature selection, and deep learning, and develop a unified machine learning potential of ab initio quality valid over a wide pressure-temperature range to determine the melting temperature of MgSiO3. The melting temperature of the high-pressure, post-perovskite phase, important for super-Earths, increases more rapidly with increasing pressure than that of the lower pressure perovskite phase, stable at the base of Earth's mantle. The volume of the liquid closely approaches that of the solid phases at the highest pressure of our study. Our computed triple point constrains the Clapeyron slope of the perovskite to post-perovskite transition, which we compare with observations of seismic reflectivity at the base of Earth's mantle to calibrate Earth's core heat flux.},
+ keywords = {/unread,AIMD,AML,compositional descriptors,DeePMD-kit,DFT,iterative learning,iterative learning scheme,LAMMPS,materials,MD,MLP,prediction of total energy,SOAP,thermodynamics,VASP},
+ file = {/Users/wasmer/Nextcloud/Zotero/Deng et al_2023_Melting of $-mathrm MgSi -mathrm O _ 3 $ determined by machine learning.pdf;/Users/wasmer/Zotero/storage/4DSIHJXI/PhysRevB.107.html}
}
@article{dennerEfficientLearningOnedimensional2020,
@@ -1713,13 +1802,12 @@
number = {3},
doi = {10.1103/PhysRevResearch.2.033388},
keywords = {DFT,ML,ML-DFT,ML-ESM,prediction of ground-state properties,topological phase transition},
- file = {/home/johannes/Nextcloud/Zotero/Denner_2020_Efficient learning of a one-dimensional density functional theory.pdf;/home/johannes/Zotero/storage/UHHVADW4/PhysRevResearch.2.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Denner_2020_Efficient learning of a one-dimensional density functional theory.pdf;/Users/wasmer/Zotero/storage/UHHVADW4/PhysRevResearch.2.html}
}
@article{depabloNewFrontiersMaterials2019,
title = {New Frontiers for the Materials Genome Initiative},
- author = {de Pablo, Juan J. and Jackson, Nicholas E. and Webb, Michael A. and Chen, Long-Qing and Moore, Joel E. and Morgan, Dane and Jacobs, Ryan and Pollock, Tresa and Schlom, Darrell G. and Toberer, Eric S. and Analytis, James and Dabo, Ismaila and DeLongchamp, Dean M. and Fiete, Gregory A. and Grason, Gregory M. and Hautier, Geoffroy and Mo, Yifei and Rajan, Krishna and Reed, Evan J. and Rodriguez, Efrain and Stevanovic, Vladan and Suntivich, Jin and Thornton, Katsuyo and Zhao, Ji-Cheng},
- options = {useprefix=true},
+ author = {family=Pablo, given=Juan J., prefix=de, useprefix=true and Jackson, Nicholas E. and Webb, Michael A. and Chen, Long-Qing and Moore, Joel E. and Morgan, Dane and Jacobs, Ryan and Pollock, Tresa and Schlom, Darrell G. and Toberer, Eric S. and Analytis, James and Dabo, Ismaila and DeLongchamp, Dean M. and Fiete, Gregory A. and Grason, Gregory M. and Hautier, Geoffroy and Mo, Yifei and Rajan, Krishna and Reed, Evan J. and Rodriguez, Efrain and Stevanovic, Vladan and Suntivich, Jin and Thornton, Katsuyo and Zhao, Ji-Cheng},
date = {2019-04-05},
journaltitle = {npj Computational Materials},
shortjournal = {npj Comput Mater},
@@ -1735,7 +1823,7 @@
issue = {1},
langid = {english},
annotation = {Bandiera\_abtest: a Cc\_license\_type: cc\_by Cg\_type: Nature Research Journals Primary\_atype: Reviews Subject\_term: Materials science;Nanoscience and technology Subject\_term\_id: materials-science;nanoscience-and-technology},
- file = {/home/johannes/Nextcloud/Zotero/de Pablo et al_2019_New frontiers for the materials genome initiative.pdf;/home/johannes/Zotero/storage/PY8DXX7D/s41524-019-0173-4.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/de Pablo et al_2019_New frontiers for the materials genome initiative.pdf;/Users/wasmer/Zotero/storage/PY8DXX7D/s41524-019-0173-4.html}
}
@article{deringerGaussianProcessRegression2021,
@@ -1754,7 +1842,27 @@
urldate = {2022-06-03},
abstract = {We provide an introduction to Gaussian process regression (GPR) machine-learning methods in computational materials science and chemistry. The focus of the present review is on the regression of atomistic properties: in particular, on the construction of interatomic potentials, or force fields, in the Gaussian Approximation Potential (GAP) framework; beyond this, we also discuss the fitting of arbitrary scalar, vectorial, and tensorial quantities. Methodological aspects of reference data generation, representation, and regression, as well as the question of how a data-driven model may be validated, are reviewed and critically discussed. A survey of applications to a variety of research questions in chemistry and materials science illustrates the rapid growth in the field. A vision is outlined for the development of the methodology in the years to come.},
keywords = {_tablet,active learning,GAP,GPR,librascal,materials,ML,ML-DFT,ML-ESM,MLP,models,molecules,prediction of electron density,prediction of LDOS,review,SA-GPR,SOAP,structure prediction,structure search,tutorial,with-code},
- file = {/home/johannes/Nextcloud/Zotero/Deringer et al_2021_Gaussian Process Regression for Materials and Molecules.pdf;/home/johannes/Zotero/storage/LSTJST2A/acs.chemrev.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Deringer et al_2021_Gaussian Process Regression for Materials and Molecules.pdf;/Users/wasmer/Zotero/storage/LSTJST2A/acs.chemrev.html}
+}
+
+@article{deringerOriginsStructuralElectronic2021,
+ title = {Origins of Structural and Electronic Transitions in Disordered Silicon},
+ author = {Deringer, Volker L. and Bernstein, Noam and Csányi, Gábor and Ben Mahmoud, Chiheb and Ceriotti, Michele and Wilson, Mark and Drabold, David A. and Elliott, Stephen R.},
+ date = {2021-01},
+ journaltitle = {Nature},
+ volume = {589},
+ number = {7840},
+ pages = {59--64},
+ publisher = {{Nature Publishing Group}},
+ issn = {1476-4687},
+ doi = {10.1038/s41586-020-03072-z},
+ url = {https://www.nature.com/articles/s41586-020-03072-z},
+ urldate = {2023-02-21},
+ abstract = {Structurally disordered materials pose fundamental questions1–4, including how different disordered phases (‘polyamorphs’) can coexist and transform from one phase to another5–9. Amorphous silicon has been extensively studied; it forms a fourfold-coordinated, covalent network at ambient conditions and much-higher-coordinated, metallic phases under pressure10–12. However, a detailed mechanistic understanding of the structural transitions in disordered silicon has been lacking, owing to the intrinsic limitations of even the most advanced experimental and computational techniques, for example, in terms of the system sizes accessible via simulation. Here we show how atomistic machine~learning models trained on accurate quantum mechanical computations can help to describe liquid–amorphous and amorphous–amorphous transitions for a system of 100,000 atoms (ten-nanometre length scale), predicting structure, stability and electronic properties. Our simulations reveal a three-step transformation sequence for amorphous silicon under increasing external pressure. First, polyamorphic low- and high-density amorphous regions are found to coexist, rather than appearing sequentially. Then, we observe a structural collapse into a distinct very-high-density amorphous (VHDA) phase. Finally, our simulations indicate the transient nature of this VHDA phase: it rapidly nucleates crystallites, ultimately leading to the formation of a polycrystalline structure, consistent with experiments13–15 but not seen in earlier simulations11,16–18. A machine~learning model for the electronic density of states confirms the onset of metallicity during VHDA formation and the subsequent crystallization. These results shed light on the liquid and amorphous states of silicon, and, in a wider context, they exemplify a machine~learning-driven approach to predictive materials modelling.},
+ issue = {7840},
+ langid = {english},
+ keywords = {/unread,delta learning,disordered,GAP,materials,MD,ML,prediction of DOS,TB,tensorial target},
+ file = {/Users/wasmer/Nextcloud/Zotero/Deringer et al_2021_Origins of structural and electronic transitions in disordered silicon.pdf}
}
@article{diceSignacDataManagement2021,
@@ -1769,7 +1877,7 @@
urldate = {2022-08-24},
eventtitle = {Proceedings of the 20th {{Python}} in {{Science Conference}}},
keywords = {Data management,RDM},
- file = {/home/johannes/Nextcloud/Zotero/Dice et al_2021_signac.pdf;/home/johannes/Zotero/storage/YML5Z3T2/bradley_dice.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Dice et al_2021_signac.pdf;/Users/wasmer/Zotero/storage/YML5Z3T2/bradley_dice.html}
}
@article{dickHighlyAccurateConstrained2021,
@@ -1787,7 +1895,7 @@
urldate = {2021-12-14},
abstract = {Using an end-to-end differentiable implementation of the Kohn-Sham self-consistent field equations, we obtain a highly accurate neural network–based exchange and correlation (XC) functional of the electronic density. The functional is optimized using information on both energy and density while exact constraints are enforced through an appropriate neural network architecture. We evaluate our model against different families of XC approximations and show that at the meta-GGA level our functional exhibits unprecedented accuracy for both energy and density predictions. For nonempirical functionals, there is a strong linear correlation between energy and density errors. We use this correlation to define an XC functional quality metric that includes both energy and density errors, leading to an improved way to rank different approximations.},
keywords = {ANN,autodiff,DFT,library,ML,ML-DFA,molecules,prediction from density,prediction of Exc,prediction of vxc,with-code},
- file = {/home/johannes/Nextcloud/Zotero/Dick_Fernandez-Serra_2021_Highly accurate and constrained density functional obtained with differentiable.pdf;/home/johannes/Zotero/storage/C5259YA7/Dick and Fernandez-Serra - 2021 - Highly accurate and constrained density functional.pdf;/home/johannes/Zotero/storage/SDPZD88A/PhysRevB.104.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Dick_Fernandez-Serra_2021_Highly accurate and constrained density functional obtained with differentiable.pdf;/Users/wasmer/Zotero/storage/C5259YA7/Dick and Fernandez-Serra - 2021 - Highly accurate and constrained density functional.pdf;/Users/wasmer/Zotero/storage/SDPZD88A/PhysRevB.104.html}
}
@article{dickLearningDensityCorrect2019,
@@ -1806,7 +1914,7 @@
urldate = {2021-12-14},
abstract = {We propose a new molecular simulation framework that combines the transferability, robustness, and chemical flexibility of an ab initio method with the accuracy and efficiency of a machine learning model. The key to achieve this mix is to use a standard density functional theory (DFT) simulation as a preprocessor for the atomic and molecular information, obtaining a good quality electronic density. General, symmetry preserving, atom-centered electronic descriptors are then built from this density to train a neural network to correct the baseline DFT energies and forces. These electronic descriptors encode much more information than local atomic environments, allowing a simple neural network to reach the accuracy required for the problem of study at a negligible additional cost. The balance between accuracy and efficiency is determined by the baseline simulation. This is shown in results where high level quantum chemical accuracy is obtained for simulations of liquid water at standard DFT cost or where high level DFT-accuracy is achieved in simulations with a low-level baseline DFT calculation at a significantly reduced cost.},
keywords = {ACSF,BPNN,DFT,ML,ML-DFA,ML-DFT,ML-ESM,MLCF,molecules,prediction from density,prediction of energy correction},
- file = {/home/johannes/Nextcloud/Zotero/Dick_Fernandez-Serra_2019_Learning from the density to correct total energy and forces in first principle.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Dick_Fernandez-Serra_2019_Learning from the density to correct total energy and forces in first principle.pdf}
}
@article{dickMachineLearningAccurate2020,
@@ -1828,7 +1936,7 @@
langid = {english},
keywords = {BPNN,DFT,library,ML,ML-DFA,MLCF,molecules,NeuralXC,prediction from density,prediction of Exc,prediction of vxc,with-code},
annotation = {Bandiera\_abtest: a Cc\_license\_type: cc\_by Cg\_type: Nature Research Journals Primary\_atype: Research Subject\_term: Density functional theory;Electronic properties and materials;Molecular dynamics Subject\_term\_id: density-functional-theory;electronic-properties-and-materials;molecular-dynamics},
- file = {/home/johannes/Nextcloud/Zotero/Dick_Fernandez-Serra_2020_Machine learning accurate exchange and correlation functionals of the.pdf;/home/johannes/Zotero/storage/95GAG2CF/s41467-020-17265-7.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Dick_Fernandez-Serra_2020_Machine learning accurate exchange and correlation functionals of the.pdf;/Users/wasmer/Zotero/storage/95GAG2CF/s41467-020-17265-7.html}
}
@article{disanteDeepLearningFunctional2022,
@@ -1846,25 +1954,23 @@
urldate = {2022-10-19},
abstract = {We perform a data-driven dimensionality reduction of the scale-dependent four-point vertex function characterizing the functional renormalization group (FRG) flow for the widely studied two-dimensional t−t′ Hubbard model on the square lattice. We demonstrate that a deep learning architecture based on a neural ordinary differential equation solver in a low-dimensional latent space efficiently learns the FRG dynamics that delineates the various magnetic and d-wave superconducting regimes of the Hubbard model. We further present a dynamic mode decomposition analysis that confirms that a small number of modes are indeed sufficient to capture the FRG dynamics. Our Letter demonstrates the possibility of using artificial intelligence to extract compact representations of the four-point vertex functions for correlated electrons, a goal of utmost importance for the success of cutting-edge quantum field theoretical methods for tackling the many-electron problem.},
keywords = {Deep learning,FRG,Hubbard model,ML,rec-by-da-silva,renormalization group,superconductor},
- file = {/home/johannes/Nextcloud/Zotero/Di Sante et al_2022_Deep Learning the Functional Renormalization Group.pdf;/home/johannes/Zotero/storage/LKT2Z79L/Di Sante et al_2022_Deep Learning the Functional Renormalization Group-supp.pdf;/home/johannes/Zotero/storage/PGSNSHSM/PhysRevLett.129.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Di Sante et al_2022_Deep Learning the Functional Renormalization Group.pdf;/Users/wasmer/Zotero/storage/LKT2Z79L/Di Sante et al_2022_Deep Learning the Functional Renormalization Group-supp.pdf;/Users/wasmer/Zotero/storage/PGSNSHSM/PhysRevLett.129.html}
}
-@misc{dominaJacobiLegendrePotential2022,
+@online{dominaJacobiLegendrePotential2022,
title = {The {{Jacobi-Legendre}} Potential},
author = {Domina, Michelangelo and Patil, Urvesh and Cobelli, Matteo and Sanvito, Stefano},
date = {2022-08-22},
number = {arXiv:2208.10292},
- eprint = {2208.10292},
+ eprint = {arXiv:2208.10292},
eprinttype = {arxiv},
- primaryclass = {cond-mat},
- publisher = {{arXiv}},
doi = {10.48550/arXiv.2208.10292},
url = {http://arxiv.org/abs/2208.10292},
urldate = {2022-09-05},
abstract = {Inspired by the cluster expansion method, we introduce a compact machine-learning potential constructed over Jacobi and Legendre polynomials. The constituent many-body contributions are separated, fully interpretable and adaptable to replicate the physical knowledge of the system, such as a repulsive behaviour at a small inter-atomic distance. Most importantly the potential requires a small number of features to achieve accuracy comparable to that of more numerically heavy and descriptor-rich alternatives. This is here tested for an organic molecule, a crystalline solid and an amorphous compound. Furthermore, we argue that the physical interpretability of the various terms is key to the selection and training of stable potentials.},
- archiveprefix = {arXiv},
- keywords = {ACE,descriptors,DFT,invariance,Jacobi-Legendre potential,JLP,linear regression,ML,ML-ESM,MLP,prediction of total energy,SNAP},
- file = {/home/johannes/Nextcloud/Zotero/Domina et al_2022_The Jacobi-Legendre potential.pdf;/home/johannes/Zotero/storage/DUUKR6TZ/2208.html}
+ pubstate = {preprint},
+ keywords = {_tablet,ACE,descriptors,DFT,invariance,Jacobi-Legendre potential,JLP,linear regression,ML,ML-ESM,MLP,prediction of total energy,SNAP},
+ file = {/Users/wasmer/Nextcloud/Zotero/Domina et al_2022_The Jacobi-Legendre potential.pdf;/Users/wasmer/Zotero/storage/DUUKR6TZ/2208.html}
}
@article{dominaSpectralNeighborRepresentation2022,
@@ -1883,7 +1989,7 @@
urldate = {2022-09-05},
abstract = {We introduce a translational and rotational invariant local representation for vector fields, which can be employed in the construction of machine learning energy models of solids and molecules. This allows us to describe, on the same footing, the energy fluctuations due to the atomic motion, the longitudinal and transverse excitations of the vector field, and their mutual interplay. The formalism can then be applied to physical systems where the total energy is determined by a vector density, as in the case of magnetism. Our representation is constructed over the power spectrum of the combined angular momentum describing the local atomic positions and the vector field, and it can be used in conjunction with different machine learning schemes and data taken from accurate ab initio electronic structure theories. We demonstrate the descriptive power of our representation for a range of classical spin Hamiltonian and machine learning algorithms. In particular, we construct energy models based on both linear Ridge regression, as in conventional spectral neighbor analysis potentials, and the Gaussian approximation. These are both built to represent a Heisenberg-type Hamiltonian including a longitudinal energy term and spin-lattice coupling.},
keywords = {_tablet,descriptors,DFT,GPR,Heisenberg model,Jij,LRR,magnetism,ML,ML-DFT,ML-ESM,spin-dependent},
- file = {/home/johannes/Nextcloud/Zotero/Domina et al_2022_Spectral neighbor representation for vector fields.pdf;/home/johannes/Zotero/storage/F4KNYWPX/Domina et al_2022_Spectral neighbor representation for vector fields.pdf;/home/johannes/Zotero/storage/QX9ZENU5/PhysRevB.105.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Domina et al_2022_Spectral neighbor representation for vector fields.pdf;/Users/wasmer/Zotero/storage/F4KNYWPX/Domina et al_2022_Spectral neighbor representation for vector fields.pdf;/Users/wasmer/Zotero/storage/QX9ZENU5/PhysRevB.105.html}
}
@unpublished{dominaSpectralneighbourRepresentationVector2022,
@@ -1893,13 +1999,117 @@
date = {2022-02-23},
eprint = {2202.13773},
eprinttype = {arxiv},
- primaryclass = {cond-mat},
+ eprintclass = {cond-mat},
url = {http://arxiv.org/abs/2202.13773},
urldate = {2022-03-23},
abstract = {We introduce a translational and rotational invariant local representation for vector fields, which can be employed in the construction of machine-learning energy models of solids and molecules. This allows us to describe, on the same footing, the energy fluctuations due to the atomic motion, the longitudinal and transverse excitations of the vector field, and their mutual interplay. The formalism can then be applied to physical systems where the total energy is determined by a vector density, as in the case of magnetism. Our representation is constructed over the power spectrum of the combined angular momentum describing the local atomic positions and the vector field, and can be used in conjunction with different machine-learning schemes and data taken from accurate ab initio electronic structure theories. We demonstrate the descriptive power of our representation for a range of classical spin Hamiltonian and machine-learning algorithms. In particular, we construct energy models based on both linear Ridge regression, as in conventional spectral neighbour analysis potentials, and gaussian approximation. These are both built to represent a Heisenberg-type Hamiltonian including a longitudinal energy term and spin-lattice coupling.},
- archiveprefix = {arXiv},
keywords = {_tablet,descriptors,DFT,GPR,Heisenberg model,Jij,LRR,magnetism,ML,ML-DFT,ML-ESM,spin-dependent},
- file = {/home/johannes/Nextcloud/Zotero/Domina et al_2022_A spectral-neighbour representation for vector fields.pdf;/home/johannes/Zotero/storage/EB6UHPCQ/2202.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Domina et al_2022_A spectral-neighbour representation for vector fields.pdf;/Users/wasmer/Zotero/storage/EB6UHPCQ/2202.html}
+}
+
+@online{dornheimMachineLearningSurrogateModel2021,
+ title = {A {{Machine-Learning Surrogate Model}} for Ab Initio {{Electronic Correlations}} at {{Extreme Conditions}}},
+ author = {Dornheim, Tobias and Moldabekov, Zhandos and Cangi, Attila},
+ date = {2021-04-07},
+ number = {arXiv:2104.02941},
+ eprint = {arXiv:2104.02941},
+ eprinttype = {arxiv},
+ doi = {10.48550/arXiv.2104.02941},
+ url = {http://arxiv.org/abs/2104.02941},
+ urldate = {2023-02-15},
+ abstract = {The electronic structure in matter under extreme conditions is a challenging complex system prevalent in astrophysical objects and highly relevant for technological applications. We show how machine-learning surrogates in terms of neural networks have a profound impact on the efficient modeling of matter under extreme conditions. We demonstrate the utility of a surrogate model that is trained on \textbackslash emph\{ab initio\} quantum Monte Carlo data for various applications in the emerging field of warm dense matter research.},
+ pubstate = {preprint},
+ keywords = {/unread},
+ file = {/Users/wasmer/Nextcloud/Zotero/Dornheim et al_2021_A Machine-Learning Surrogate Model for ab initio Electronic Correlations at.pdf;/Users/wasmer/Zotero/storage/F4428MJB/2104.html}
+}
+
+@article{dragoniAchievingDFTAccuracy2018,
+ title = {Achieving {{DFT}} Accuracy with a Machine-Learning Interatomic Potential: {{Thermomechanics}} and Defects in Bcc Ferromagnetic Iron},
+ shorttitle = {Achieving {{DFT}} Accuracy with a Machine-Learning Interatomic Potential},
+ author = {Dragoni, Daniele and Daff, Thomas D. and Csányi, Gábor and Marzari, Nicola},
+ date = {2018-01-30},
+ journaltitle = {Physical Review Materials},
+ shortjournal = {Phys. Rev. Mater.},
+ volume = {2},
+ number = {1},
+ pages = {013808},
+ publisher = {{American Physical Society}},
+ doi = {10.1103/PhysRevMaterials.2.013808},
+ url = {https://link.aps.org/doi/10.1103/PhysRevMaterials.2.013808},
+ urldate = {2023-03-12},
+ abstract = {We show that the Gaussian Approximation Potential (GAP) machine-learning framework can describe complex magnetic potential energy surfaces, taking ferromagnetic iron as a paradigmatic challenging case. The training database includes total energies, forces, and stresses obtained from density-functional theory in the generalized-gradient approximation, and comprises approximately 150,000 local atomic environments, ranging from pristine and defected bulk configurations to surfaces and generalized stacking faults with different crystallographic orientations. We find the structural, vibrational, and thermodynamic properties of the GAP model to be in excellent agreement with those obtained directly from first-principles electronic-structure calculations. There is good transferability to quantities, such as Peierls energy barriers, which are determined to a large extent by atomic configurations that were not part of the training set. We observe the benefit and the need of using highly converged electronic-structure calculations to sample a target potential energy surface. The end result is a systematically improvable potential that can achieve the same accuracy of density-functional theory calculations, but at a fraction of the computational cost.},
+ keywords = {/unread,AML,bulk defects,collinear,CUR decomposition,database generation,defects,DFT,disordered,Ferromagnetism,GAP,GGA,GPR,interstitials,LAMMPS,line defects,materials,MD,ML,PBE,point defects,Quantum ESPRESSO,QUIP,SOAP,spin-polarized,vacancies},
+ file = {/Users/wasmer/Nextcloud/Zotero/Dragoni et al_2018_Achieving DFT accuracy with a machine-learning interatomic potential.pdf;/Users/wasmer/Zotero/storage/H8ISZZP6/PhysRevMaterials.2.html}
+}
+
+@online{dralBookQuantumChemistry2022,
+ title = {Book “{{Quantum Chemistry}} in the {{Age}} of {{Machine Learning}}â€},
+ author = {Dral, Pavlo},
+ date = {2022-09-20T09:27:08+00:00},
+ url = {http://dr-dral.com/book-quantum-chemistry-in-the-age-of-machine-learning/},
+ urldate = {2023-03-02},
+ abstract = {The book “Quantum Chemistry in the Age of Machine Learning†guides aspiring beginners and specialists in this exciting field by covering topics ranging from basic concepts to comprehens…},
+ langid = {american},
+ organization = {{Dral's Group}},
+ keywords = {/unread,AML,book,ML,ML-ESM,prediction of electron density,tutorial,with-code},
+ file = {/Users/wasmer/Zotero/storage/ZPMKDHTP/book-quantum-chemistry-in-the-age-of-machine-learning.html}
+}
+
+@article{dralMLatomIntegrativePlatform2021,
+ title = {{{MLatom}} 2: {{An Integrative Platform}} for {{Atomistic Machine Learning}}},
+ shorttitle = {{{MLatom}} 2},
+ author = {Dral, Pavlo O. and Ge, Fuchun and Xue, Bao-Xin and Hou, Yi-Fan and Pinheiro, Max and Huang, Jianxing and Barbatti, Mario},
+ date = {2021-06-08},
+ journaltitle = {Topics in Current Chemistry},
+ shortjournal = {Top Curr Chem (Z)},
+ volume = {379},
+ number = {4},
+ pages = {27},
+ issn = {2364-8961},
+ doi = {10.1007/s41061-021-00339-5},
+ url = {https://doi.org/10.1007/s41061-021-00339-5},
+ urldate = {2023-03-02},
+ abstract = {Atomistic machine learning (AML) simulations are used in chemistry at an ever-increasing pace. A large number of AML models has been developed, but their implementations are scattered among different packages, each with its own conventions for input and output. Thus, here we give an overview of our MLatom 2 software package, which provides an integrative platform for a wide variety of AML simulations by implementing from scratch and interfacing existing software for a range of state-of-the-art models. These include kernel method-based model types such as KREG (native implementation), sGDML, and GAP-SOAP as well as neural-network-based model types such as ANI, DeepPot-SE, and PhysNet. The theoretical foundations behind these methods are overviewed too. The modular structure of MLatom allows for easy extension to more AML model types. MLatom 2 also has many other capabilities useful for AML simulations, such as the support of custom descriptors, farthest-point and structure-based sampling, hyperparameter optimization, model evaluation, and automatic learning curve generation. It can also be used for such multi-step tasks as Δ-learning, self-correction approaches, and absorption spectrum simulation within the machine-learning nuclear-ensemble approach. Several of these MLatom 2 capabilities are showcased in application examples.},
+ langid = {english},
+ keywords = {/unread,AML,benchmarking,GPR,KRR,library,ML,MLatom,MLP,NN,Python},
+ file = {/Users/wasmer/Nextcloud/Zotero/Dral et al_2021_MLatom 2.pdf}
+}
+
+@article{dralMLatomProgramPackage2019,
+ title = {{{MLatom}}: {{A}} Program Package for Quantum Chemical Research Assisted by Machine Learning},
+ shorttitle = {{{MLatom}}},
+ author = {Dral, Pavlo O.},
+ date = {2019},
+ journaltitle = {Journal of Computational Chemistry},
+ volume = {40},
+ number = {26},
+ pages = {2339--2347},
+ issn = {1096-987X},
+ doi = {10.1002/jcc.26004},
+ url = {https://onlinelibrary.wiley.com/doi/abs/10.1002/jcc.26004},
+ urldate = {2023-03-02},
+ abstract = {MLatom is a program package designed for computationally efficient simulations of atomistic systems with machine-learning (ML) algorithms. It can be used out-of-the-box as a stand-alone program with a user-friendly online manual. The use of MLatom does not require extensive knowledge of machine learning, programming, or scripting. The user need only prepare input files and choose appropriate options. The program implements kernel ridge regression and supports Gaussian, Laplacian, and Matérn kernels. It can use arbitrary, user-provided input vectors and can convert molecular geometries into input vectors corresponding to several types of built-in molecular descriptors. MLatom saves and re-uses trained ML models as needed, in addition to estimating the generalization error of ML setups. Various sampling procedures are supported and the gradients of output properties can be calculated. The core part of MLatom is written in Fortran, uses standard libraries for linear algebra, and is optimized for shared-memory parallel computations. © 2019 Wiley Periodicals, Inc.},
+ langid = {english},
+ keywords = {/unread,AML,CM,descriptors,kernel methods,KRR,library,MLatom,MLP,NN,original publication,PhysNet,Python,sampling,SchNet,sGDML,with-code},
+ file = {/Users/wasmer/Nextcloud/Zotero/Dral_2019_MLatom.pdf;/Users/wasmer/Zotero/storage/4MWX9JQP/jcc.html}
+}
+
+@article{dralQuantumChemistryAge2020,
+ title = {Quantum {{Chemistry}} in the {{Age}} of {{Machine Learning}}},
+ author = {Dral, Pavlo O.},
+ date = {2020-03-19},
+ journaltitle = {The Journal of Physical Chemistry Letters},
+ shortjournal = {J. Phys. Chem. Lett.},
+ volume = {11},
+ number = {6},
+ pages = {2336--2347},
+ publisher = {{American Chemical Society}},
+ doi = {10.1021/acs.jpclett.9b03664},
+ url = {https://doi.org/10.1021/acs.jpclett.9b03664},
+ urldate = {2023-03-02},
+ abstract = {As the quantum chemistry (QC) community embraces machine learning (ML), the number of new methods and applications based on the combination of QC and ML is surging. In this Perspective, a view of the current state of affairs in this new and exciting research field is offered, challenges of using machine learning in quantum chemistry applications are described, and potential future developments are outlined. Specifically, examples of how machine learning is used to improve the accuracy and accelerate quantum chemical research are shown. Generalization and classification of existing techniques are provided to ease the navigation in the sea of literature and to guide researchers entering the field. The emphasis of this Perspective is on supervised machine learning.},
+ keywords = {/unread,AML,delta learning,GPR,KRR,ML,ML-DFT,ML-ESM,ML-FF,MLatom,MLP,NN,Quantum chemistry,review},
+ file = {/Users/wasmer/Nextcloud/Zotero/Dral_2020_Quantum Chemistry in the Age of Machine Learning.pdf;/Users/wasmer/Zotero/storage/4NY56BJV/acs.jpclett.html}
}
@article{drautzAtomicClusterExpansion2019,
@@ -1917,7 +2127,7 @@
urldate = {2022-05-11},
abstract = {The atomic cluster expansion is developed as a complete descriptor of the local atomic environment, including multicomponent materials, and its relation to a number of other descriptors and potentials is discussed. The effort for evaluating the atomic cluster expansion is shown to scale linearly with the number of neighbors, irrespective of the order of the expansion. Application to small Cu clusters demonstrates smooth convergence of the atomic cluster expansion to meV accuracy. By introducing nonlinear functions of the atomic cluster expansion an interatomic potential is obtained that is comparable in accuracy to state-of-the-art machine learning potentials. Because of the efficient convergence of the atomic cluster expansion relevant subspaces can be sampled uniformly and exhaustively. This is demonstrated by testing against a large database of density functional theory calculations for copper.},
keywords = {_tablet,ACE,descriptors,original publication},
- file = {/home/johannes/Nextcloud/Zotero/Drautz_2019_Atomic cluster expansion for accurate and transferable interatomic potentials.pdf;/home/johannes/Zotero/storage/HNR9ZCLL/Drautz_2019_Atomic cluster expansion for accurate and transferable interatomic potentials.pdf;/home/johannes/Zotero/storage/NMAUF3NJ/PhysRevB.99.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Drautz_2019_Atomic cluster expansion for accurate and transferable interatomic potentials.pdf;/Users/wasmer/Zotero/storage/HNR9ZCLL/Drautz_2019_Atomic cluster expansion for accurate and transferable interatomic potentials.pdf;/Users/wasmer/Zotero/storage/NMAUF3NJ/PhysRevB.99.html}
}
@article{drautzAtomicClusterExpansion2020,
@@ -1930,25 +2140,23 @@
number = {2},
doi = {10.1103/PhysRevB.102.024104},
keywords = {_tablet,ACE,descriptors,magnetism,ML,spin-dependent},
- file = {/home/johannes/Nextcloud/Zotero/Drautz_2020_Atomic cluster expansion of scalar, vectorial, and tensorial properties.pdf;/home/johannes/Zotero/storage/9W2WE4WX/PhysRevB.102.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Drautz_2020_Atomic cluster expansion of scalar, vectorial, and tensorial properties.pdf;/Users/wasmer/Zotero/storage/9W2WE4WX/PhysRevB.102.html}
}
-@misc{drautzAtomicClusterExpansion2022,
+@online{drautzAtomicClusterExpansion2022,
title = {Atomic Cluster Expansion and Wave Function Representations},
author = {Drautz, Ralf and Ortner, Christoph},
date = {2022-06-22},
number = {arXiv:2206.11375},
- eprint = {2206.11375},
+ eprint = {arXiv:2206.11375},
eprinttype = {arxiv},
- primaryclass = {cond-mat},
- publisher = {{arXiv}},
doi = {10.48550/arXiv.2206.11375},
url = {http://arxiv.org/abs/2206.11375},
urldate = {2022-06-28},
abstract = {The atomic cluster expansion (ACE) has been highly successful for the parameterisation of symmetric (invariant or equivariant) properties of many-particle systems. Here, we generalize its derivation to anti-symmetric functions. We show how numerous well-known linear representations of wave functions naturally arise within this framework and we explore how recent successful nonlinear parameterisations can be further enhanced by employing ACE methodology. From this analysis we propose a wide design space of promising wave function representations.},
- archiveprefix = {arXiv},
+ pubstate = {preprint},
keywords = {_tablet,ACE,Backflow,Deep learning,ML-QM,prediction of wavefunction,representation of wavefunction,Slater-Jastrow},
- file = {/home/johannes/Nextcloud/Zotero/Drautz_Ortner_2022_Atomic cluster expansion and wave function representations.pdf;/home/johannes/Zotero/storage/6PTQT7NH/2206.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Drautz_Ortner_2022_Atomic cluster expansion and wave function representations.pdf;/Users/wasmer/Zotero/storage/6PTQT7NH/2206.html}
}
@article{drautzSpinclusterExpansionParametrization2004,
@@ -1962,7 +2170,7 @@
number = {10},
doi = {10.1103/PhysRevB.69.104404},
keywords = {ACE-related,descriptors},
- file = {/home/johannes/Zotero/storage/IICKDCFM/PhysRevB.69.html}
+ file = {/Users/wasmer/Zotero/storage/IICKDCFM/PhysRevB.69.html}
}
@article{draxlNOMADFAIRConcept2018,
@@ -1982,7 +2190,7 @@
abstract = {, Data are a crucial raw material of this century. The amount of data that have been created in materials science thus far and that continues to be created every day is immense. Without a proper infrastructure that allows for collecting and sharing data, the envisioned success of big data-driven materials science will be hampered. For the field of computational materials science, the NOMAD (Novel Materials Discovery) Center of Excellence (CoE) has changed the scientific culture toward comprehensive and findable, accessible, interoperable, and reusable (FAIR) data, opening new avenues for mining materials science big data. Novel data-analytics concepts and tools turn data into knowledge and help in the prediction of new materials and in the identification of new properties of already known materials.},
langid = {english},
keywords = {artificial intelligence,data repositories,data sharing,machine learning,metadata},
- file = {/home/johannes/Nextcloud/Zotero/Draxl_Scheffler_2018_NOMAD.pdf;/home/johannes/Zotero/storage/3W2KJMWA/1EEF321F62D41997CA16AD367B74C4B0.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Draxl_Scheffler_2018_NOMAD.pdf;/Users/wasmer/Zotero/storage/3W2KJMWA/1EEF321F62D41997CA16AD367B74C4B0.html}
}
@article{draxlNOMADLaboratoryData2019,
@@ -2001,7 +2209,7 @@
urldate = {2021-10-15},
abstract = {The Novel Materials Discovery (NOMAD) Laboratory is a user-driven platform for sharing and exploiting computational materials science data. It accounts for the various aspects of data being a crucial raw material and most relevant to accelerate materials research and engineering. NOMAD, with the NOMAD Repository, and its code-independent and normalized form, the NOMAD Archive, comprises the worldwide largest data collection of this field. Based on its findable accessible, interoperable, reusable data infrastructure, various services are offered, comprising advanced visualization, the NOMAD Encyclopedia, and artificial-intelligence tools. The latter are realized in the NOMAD Analytics Toolkit. Prerequisite for all this is the NOMAD metadata, a unique and thorough description of the data, that are produced by all important computer codes of the community. Uploaded data are tagged by a persistent identifier, and users can also request a digital object identifier to make data citable. Developments and advancements of parsers and metadata are organized jointly with users and code developers. In this work, we review the NOMAD concept and implementation, highlight its orthogonality to and synergistic interplay with other data collections, and provide an outlook regarding ongoing and future developments.},
langid = {english},
- file = {/home/johannes/Nextcloud/Zotero/Draxl_Scheffler_2019_The NOMAD laboratory.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Draxl_Scheffler_2019_The NOMAD laboratory.pdf}
}
@book{dresselhausGroupTheory2007,
@@ -2014,8 +2222,8 @@
abstract = {Application to the Physics of Condensed Matter},
isbn = {978-3-540-32897-1},
langid = {english},
- keywords = {condensed matter,group theory,irreps,learning material,mathematics,rec-by-sabastian,textbook},
- file = {/home/johannes/Nextcloud/Zotero/Dresselhaus et al. - 2007 - Group Theory.pdf;/home/johannes/Zotero/storage/GGGVNLC4/978-3-540-32899-5.html}
+ keywords = {_tablet,condensed matter,group theory,irreps,learning material,mathematics,rec-by-sabastian,textbook},
+ file = {/Users/wasmer/Nextcloud/Zotero/Dresselhaus et al_2007_Group Theory.pdf;/Users/wasmer/Zotero/storage/GGGVNLC4/978-3-540-32899-5.html}
}
@article{drozdovConventionalSuperconductivity2032015,
@@ -2036,7 +2244,7 @@
langid = {english},
keywords = {applications of DFT,DFT,master-thesis,superconductor},
annotation = {Bandiera\_abtest: a Cg\_type: Nature Research Journals Primary\_atype: Research Subject\_term: Superconducting properties and materials Subject\_term\_id: superconducting-properties-and-materials},
- file = {/home/johannes/Nextcloud/Zotero/Drozdov et al_2015_Conventional superconductivity at 203 kelvin at high pressures in the sulfur.pdf;/home/johannes/Zotero/storage/CJIZLLVA/nature14964.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Drozdov et al_2015_Conventional superconductivity at 203 kelvin at high pressures in the sulfur.pdf;/Users/wasmer/Zotero/storage/CJIZLLVA/nature14964.html}
}
@article{dunnBenchmarkingMaterialsProperty2020,
@@ -2058,23 +2266,21 @@
issue = {1},
langid = {english},
annotation = {Bandiera\_abtest: a Cc\_license\_type: cc\_by Cg\_type: Nature Research Journals Primary\_atype: Research Subject\_term: Atomistic models;Computational methods Subject\_term\_id: atomistic-models;computational-methods},
- file = {/home/johannes/Nextcloud/Zotero/Dunn et al_2020_Benchmarking materials property prediction methods.pdf;/home/johannes/Zotero/storage/N76WQWKL/s41524-020-00406-3.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Dunn et al_2020_Benchmarking materials property prediction methods.pdf;/Users/wasmer/Zotero/storage/N76WQWKL/s41524-020-00406-3.html}
}
@unpublished{dussonAtomicClusterExpansion2021,
title = {Atomic {{Cluster Expansion}}: {{Completeness}}, {{Efficiency}} and {{Stability}}},
shorttitle = {Atomic {{Cluster Expansion}}},
- author = {Dusson, Genevieve and Bachmayr, Markus and Csanyi, Gabor and Drautz, Ralf and Etter, Simon and van der Oord, Cas and Ortner, Christoph},
- options = {useprefix=true},
+ author = {Dusson, Genevieve and Bachmayr, Markus and Csanyi, Gabor and Drautz, Ralf and Etter, Simon and family=Oord, given=Cas, prefix=van der, useprefix=true and Ortner, Christoph},
date = {2021-05-12},
eprint = {1911.03550},
eprinttype = {arxiv},
- primaryclass = {cs, math},
+ eprintclass = {cs, math},
url = {http://arxiv.org/abs/1911.03550},
urldate = {2022-05-11},
abstract = {The Atomic Cluster Expansion (Drautz, Phys. Rev. B 99, 2019) provides a framework to systematically derive polynomial basis functions for approximating isometry and permutation invariant functions, particularly with an eye to modelling properties of atomistic systems. Our presentation extends the derivation by proposing a precomputation algorithm that yields immediate guarantees that a complete basis is obtained. We provide a fast recursive algorithm for efficient evaluation and illustrate its performance in numerical tests. Finally, we discuss generalisations and open challenges, particularly from a numerical stability perspective, around basis optimisation and parameter estimation, paving the way towards a comprehensive analysis of the convergence to a high-fidelity reference model.},
- archiveprefix = {arXiv},
- file = {/home/johannes/Nextcloud/Zotero/Dusson et al_2021_Atomic Cluster Expansion.pdf;/home/johannes/Zotero/storage/7WDUQE6K/1911.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Dusson et al_2021_Atomic Cluster Expansion.pdf;/Users/wasmer/Zotero/storage/7WDUQE6K/1911.html}
}
@unpublished{dymLowDimensionalInvariant2022,
@@ -2084,15 +2290,14 @@
number = {arXiv:2205.02956},
eprint = {2205.02956},
eprinttype = {arxiv},
- primaryclass = {cs, math},
+ eprintclass = {cs, math},
publisher = {{arXiv}},
doi = {10.48550/arXiv.2205.02956},
url = {http://arxiv.org/abs/2205.02956},
urldate = {2022-05-18},
abstract = {This paper studies separating invariants: mappings on \$d\$-dimensional semi-algebraic subsets of \$D\$ dimensional Euclidean domains which are invariant to semi-algebraic group actions and separate orbits. The motivation for this study comes from the usefulness of separating invariants in proving universality of equivariant neural network architectures. We observe that in several cases the cardinality of separating invariants proposed in the machine learning literature is much larger than the ambient dimension \$D\$. As a result, the theoretical universal constructions based on these separating invariants is unrealistically large. Our goal in this paper is to resolve this issue. We show that when a continuous family of semi-algebraic separating invariants is available, separation can be obtained by randomly selecting \$2d+1 \$ of these invariants. We apply this methodology to obtain an efficient scheme for computing separating invariants for several classical group actions which have been studied in the invariant learning literature. Examples include matrix multiplication actions on point clouds by permutations, rotations, and various other linear groups.},
- archiveprefix = {arXiv},
keywords = {geometric deep learning,invariance},
- file = {/home/johannes/Nextcloud/Zotero/Dym_Gortler_2022_Low Dimensional Invariant Embeddings for Universal Geometric Learning.pdf;/home/johannes/Zotero/storage/I8BIG3VX/2205.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Dym_Gortler_2022_Low Dimensional Invariant Embeddings for Universal Geometric Learning.pdf;/Users/wasmer/Zotero/storage/I8BIG3VX/2205.html}
}
@article{ebertCalculatingCondensedMatter2011,
@@ -2112,7 +2317,7 @@
abstract = {The modern version of the KKR (Korringa–Kohn–Rostoker) method represents the electronic structure of a system directly and efficiently in terms of its single-particle Green's function (GF). This is in contrast to its original version and many other traditional wave-function-based all-electron band structure methods dealing with periodically ordered solids. Direct access to the GF results in several appealing features. In addition, a wide applicability of the method is achieved by employing multiple scattering theory. The basic ideas behind the resulting KKR-GF method are outlined and the different techniques to deal with the underlying multiple scattering problem are reviewed. Furthermore, various applications of the KKR-GF method are reviewed in some detail to demonstrate the remarkable flexibility of the approach. Special attention is devoted to the numerous developments of the KKR-GF method, that have been contributed in recent years by a number of work groups, in particular in the following fields: embedding schemes for atoms, clusters and surfaces, magnetic response functions and anisotropy, electronic and spin-dependent transport, dynamical mean field theory, various kinds of spectroscopies, as well as first-principles determination of model parameters.},
langid = {english},
keywords = {_tablet,KKR,review},
- file = {/home/johannes/Nextcloud/Zotero/Ebert et al_2011_Calculating condensed matter properties using the KKR-Green's function.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Ebert et al_2011_Calculating condensed matter properties using the KKR-Green's function.pdf}
}
@unpublished{eckhoffHighDimensionalNeuralNetwork2021,
@@ -2121,13 +2326,12 @@
date = {2021-04-29},
eprint = {2104.14439},
eprinttype = {arxiv},
- primaryclass = {physics},
+ eprintclass = {physics},
url = {http://arxiv.org/abs/2104.14439},
urldate = {2021-05-18},
abstract = {Machine learning potentials have emerged as a powerful tool to extend the time and length scales of first principles-quality simulations. Still, most machine learning potentials cannot distinguish different electronic spin orientations and thus are not applicable to materials in different magnetic states. Here, we propose spin-dependent atom-centered symmetry functions as a new type of descriptor taking the atomic spin degrees of freedom into account. When used as input for a high-dimensional neural network potential (HDNNP), accurate potential energy surfaces of multicomponent systems describing multiple magnetic states can be constructed. We demonstrate the performance of these magnetic HDNNPs for the case of manganese oxide, MnO. We show that the method predicts the magnetically distorted rhombohedral structure in excellent agreement with density functional theory and experiment. Its efficiency allows to determine the N\textbackslash '\{e\}el temperature considering structural fluctuations, entropic effects, and defects. The method is general and is expected to be useful also for other types of systems like oligonuclear transition metal complexes.},
- archiveprefix = {arXiv},
keywords = {ACSF,ANN,descriptors,HDNNP,Heisenberg model,magnetism,ML,MLP,models,Physics - Computational Physics,spin-dependent,to read 2105},
- file = {/home/johannes/Nextcloud/Zotero/Eckhoff_Behler_2021_High-Dimensional Neural Network Potentials for Magnetic Systems Using.pdf;/home/johannes/Zotero/storage/KW8NBSDW/2104.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Eckhoff_Behler_2021_High-Dimensional Neural Network Potentials for Magnetic Systems Using.pdf;/Users/wasmer/Zotero/storage/KW8NBSDW/2104.html}
}
@article{ellisAcceleratingFinitetemperatureKohnSham2021,
@@ -2145,7 +2349,7 @@
urldate = {2021-12-05},
abstract = {We present a numerical modeling workflow based on machine learning which reproduces the total energies produced by Kohn-Sham density functional theory (DFT) at finite electronic temperature to within chemical accuracy at negligible computational cost. Based on deep neural networks, our workflow yields the local density of states (LDOS) for a given atomic configuration. From the LDOS, spatially resolved, energy-resolved, and integrated quantities can be calculated, including the DFT total free energy, which serves as the Born-Oppenheimer potential energy surface for the atoms. We demonstrate the efficacy of this approach for both solid and liquid metals and compare results between independent and unified machine-learning models for solid and liquid aluminum. Our machine-learning density functional theory framework opens up the path towards multiscale materials modeling for matter under ambient and extreme conditions at a computational scale and cost that is unattainable with current algorithms.},
keywords = {DFT,finite-temperature DFT,LAMMPS,library,MALA,ML,ML-DFT,ML-ESM,prediction of LDOS,quantum,Quantum ESPRESSO,SNAP,with-code},
- file = {/home/johannes/Nextcloud/Zotero/Ellis et al_2021_Accelerating finite-temperature Kohn-Sham density functional theory with deep.pdf;/home/johannes/Zotero/storage/AS2E35V9/PhysRevB.104.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Ellis et al_2021_Accelerating finite-temperature Kohn-Sham density functional theory with deep.pdf;/Users/wasmer/Zotero/storage/AS2E35V9/PhysRevB.104.html}
}
@article{evansGroupTheory2004,
@@ -2156,12 +2360,12 @@
urldate = {2022-12-05},
abstract = {Group Theory},
keywords = {condensed matter,group theory,irreps,learning material,lecture notes,mathematics,rec-by-sabastian,textbook},
- file = {/home/johannes/Nextcloud/Zotero/Evans_2004_Group Theory.pdf;/home/johannes/Zotero/storage/K7KNP9VQ/Group_Theory.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Evans_2004_Group Theory.pdf;/Users/wasmer/Zotero/storage/K7KNP9VQ/Group_Theory.html}
}
@article{faberCrystalStructureRepresentations2015,
title = {Crystal Structure Representations for Machine Learning Models of Formation Energies},
- author = {Faber, Felix and Lindmaa, Alexander and von Lilienfeld, O. Anatole and Armiento, Rickard},
+ author = {Faber, Felix and Lindmaa, Alexander and family=Lilienfeld, given=O. Anatole, prefix=von, useprefix=false and Armiento, Rickard},
date = {2015},
journaltitle = {International Journal of Quantum Chemistry},
volume = {115},
@@ -2174,8 +2378,7 @@
abstract = {We introduce and evaluate a set of feature vector representations of crystal structures for machine learning (ML) models of formation energies of solids. ML models of atomization energies of organic molecules have been successful using a Coulomb matrix representation of the molecule. We consider three ways to generalize such representations to periodic systems: (i) a matrix where each element is related to the Ewald sum of the electrostatic interaction between two different atoms in the unit cell repeated over the lattice; (ii) an extended Coulomb-like matrix that takes into account a number of neighboring unit cells; and (iii) an ansatz that mimics the periodicity and the basic features of the elements in the Ewald sum matrix using a sine function of the crystal coordinates of the atoms. The representations are compared for a Laplacian kernel with Manhattan norm, trained to reproduce formation energies using a dataset of 3938 crystal structures obtained from the Materials Project. For training sets consisting of 3000 crystals, the generalization error in predicting formation energies of new structures corresponds to (i) 0.49, (ii) 0.64, and (iii) for the respective representations. © 2015 Wiley Periodicals, Inc.},
langid = {english},
keywords = {descriptors,Ewald sum matrix,ML,original publication,Sine matrix},
- annotation = {\_eprint: https://onlinelibrary.wiley.com/doi/pdf/10.1002/qua.24917},
- file = {/home/johannes/Nextcloud/Zotero/Faber et al_2015_Crystal structure representations for machine learning models of formation.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Faber et al_2015_Crystal structure representations for machine learning models of formation.pdf}
}
@article{fabianiInvestigatingUltrafastQuantum2019,
@@ -2193,7 +2396,7 @@
abstract = {SciPost Journals Publication Detail SciPost Phys. 7, 004 (2019) Investigating ultrafast quantum magnetism with machine learning},
langid = {english},
keywords = {Heisenberg model,magnetism,ML,NN,RBM,rec-by-bluegel,spin-dependent},
- file = {/home/johannes/Nextcloud/Zotero/Fabiani_Mentink_2019_Investigating ultrafast quantum magnetism with machine learning.pdf;/home/johannes/Zotero/storage/U8VR8E9L/SciPostPhys.7.1.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Fabiani_Mentink_2019_Investigating ultrafast quantum magnetism with machine learning.pdf;/Users/wasmer/Zotero/storage/U8VR8E9L/SciPostPhys.7.1.html}
}
@article{fabianiSupermagnonicPropagationTwoDimensional2021,
@@ -2211,7 +2414,7 @@
urldate = {2022-03-29},
abstract = {We investigate the propagation of magnons after ultrashort perturbations of the exchange interaction in the prototype two-dimensional Heisenberg antiferromagnet. Using the recently proposed neural quantum states, we predict highly anisotropic spreading in space constrained by the symmetry of the perturbation. Interestingly, the propagation speed at the shortest length scale and timescale is up to 40\% higher than the highest magnon velocity. We argue that the enhancement stems from extraordinary strong magnon-magnon interactions, suggesting new avenues for manipulating information transfer on ultrashort length scales and timescales.},
keywords = {rec-by-bluegel},
- file = {/home/johannes/Nextcloud/Zotero/Fabiani et al_2021_Supermagnonic Propagation in Two-Dimensional Antiferromagnets.pdf;/home/johannes/Zotero/storage/VBV4L6RZ/Fabiani et al. - 2021 - Supermagnonic Propagation in Two-Dimensional Antif.pdf;/home/johannes/Zotero/storage/ZLEKQ276/PhysRevLett.127.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Fabiani et al_2021_Supermagnonic Propagation in Two-Dimensional Antiferromagnets.pdf;/Users/wasmer/Zotero/storage/VBV4L6RZ/Fabiani et al. - 2021 - Supermagnonic Propagation in Two-Dimensional Antif.pdf;/Users/wasmer/Zotero/storage/ZLEKQ276/PhysRevLett.127.html}
}
@article{fabrizioElectronDensityLearning2019,
@@ -2228,7 +2431,7 @@
urldate = {2021-10-16},
langid = {english},
keywords = {ML,ML-DFT,ML-ESM,prediction of electron density},
- file = {/home/johannes/Nextcloud/Zotero/Fabrizio et al_2019_Electron density learning of non-covalent systems.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Fabrizio et al_2019_Electron density learning of non-covalent systems.pdf}
}
@article{farajiHighAccuracyTransferability2017,
@@ -2246,7 +2449,7 @@
urldate = {2021-07-22},
abstract = {We investigate the accuracy and transferability of a recently developed high-dimensional neural network (NN) method for calcium fluoride, fitted to a database of ab initio density functional theory (DFT) calculations based on the Perdew-Burke-Ernzerhof (PBE) exchange correlation functional. We call the method charge equilibration via neural network technique (CENT). Although the fitting database contains only clusters (i.e., nonperiodic structures), the NN scheme accurately describes a variety of bulk properties. In contrast to other available empirical methods the CENT potential has a much simpler functional form, nevertheless it correctly reproduces the PBE energetics of various crystalline phases both at ambient and high pressure. Surface energies and structures as well as dynamical properties derived from phonon calculations are also in good agreement with PBE results. Overall, the difference between the values obtained by the CENT potential and the PBE reference values is less than or equal to the difference between the values of local density approximation (LDA) and Born-Mayer-Huggins (BMH) with those calculated by the PBE exchange correlation functional.},
keywords = {CENT,HDNNP,rec-by-bluegel},
- file = {/home/johannes/Nextcloud/Zotero/Faraji et al_2017_High accuracy and transferability of a neural network potential through charge.pdf;/home/johannes/Zotero/storage/GU7MU2BP/PhysRevB.95.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Faraji et al_2017_High accuracy and transferability of a neural network potential through charge.pdf;/Users/wasmer/Zotero/storage/GU7MU2BP/PhysRevB.95.html}
}
@article{fernandez-delgadoWeNeedHundreds2014,
@@ -2261,7 +2464,7 @@
url = {http://jmlr.org/papers/v15/delgado14a.html},
urldate = {2022-01-02},
abstract = {We evaluate 179 classifiers arising from 17 families (discriminant analysis, Bayesian, neural networks, support vector machines, decision trees, rule-based classifiers, boosting, bagging, stacking, random forests and other ensembles, generalized linear models, nearest-neighbors, partial least squares and principal component regression, logistic and multinomial regression, multiple adaptive regression splines and other methods), implemented in Weka, R (with and without the caret package), C and Matlab, including all the relevant classifiers available today. We use 121 data sets, which represent the whole UCI data base (excluding the large- scale problems) and other own real problems, in order to achieve significant conclusions about the classifier behavior, not dependent on the data set collection. The classifiers most likely to be the bests are the random forest (RF) versions, the best of which (implemented in R and accessed via caret) achieves 94.1\% of the maximum accuracy overcoming 90\% in the 84.3\% of the data sets. However, the difference is not statistically significant with the second best, the SVM with Gaussian kernel implemented in C using LibSVM, which achieves 92.3\% of the maximum accuracy. A few models are clearly better than the remaining ones: random forest, SVM with Gaussian and polynomial kernels, extreme learning machine with Gaussian kernel, C5.0 and avNNet (a committee of multi-layer perceptrons implemented in R with the caret package). The random forest is clearly the best family of classifiers (3 out of 5 bests classifiers are RF), followed by SVM (4 classifiers in the top-10), neural networks and boosting ensembles (5 and 3 members in the top-20, respectively).},
- file = {/home/johannes/Nextcloud/Zotero/Fernández-Delgado et al_2014_Do we Need Hundreds of Classifiers to Solve Real World Classification Problems.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Fernández-Delgado et al_2014_Do we Need Hundreds of Classifiers to Solve Real World Classification Problems.pdf}
}
@article{feyPowerMessagePassing2022,
@@ -2273,9 +2476,25 @@
urldate = {2022-10-16},
abstract = {This thesis proposes novel approaches for machine learning on irregularly structured input data such as graphs, point clouds and manifolds. Specifically, we are breaking up with the regularity restriction of conventional deep learning techniques, and propose solutions in designing, implementing and scaling up deep end-to-end representation learning on graph-structured data, known as Graph Neural Networks (GNNs). GNNs capture local graph structure and feature information by following a neural message passing scheme, in which node representations are recursively updated in a trainable and purely local fashion. In this thesis, we demonstrate the generality of message passing through a unified framework suitable for a wide range of operators and learning tasks. Specifically, we analyze the limitations and inherent weaknesses of GNNs and propose efficient solutions to overcome them, both theoretically and in practice, e.g., by conditioning messages via continuous B-spline kernels, by utilizing hierarchical message passing, or by leveraging positional encodings. In addition, we ensure that our proposed methods scale naturally to large input domains. In particular, we propose novel methods to fully eliminate the exponentially increasing dependency of nodes over layers inherent to message passing GNNs. Lastly, we introduce PyTorch Geometric, a deep learning library for implementing and working with graph-based neural network building blocks, built upon PyTorch.},
langid = {english},
- keywords = {GDL,GNN,library,MPNN,PhD,PyG,pytorch,thesis},
+ keywords = {GDL,GNN,library,MPNN,original publication,PhD,PyG,pytorch,thesis,with-code},
annotation = {Accepted: 2022-08-31T09:01:01Z},
- file = {/home/johannes/Nextcloud/Zotero/Fey_2022_On the power of message passing for learning on graph-structured data.pdf;/home/johannes/Zotero/storage/CEBUDJT7/41059.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Fey_2022_On the power of message passing for learning on graph-structured data.pdf;/Users/wasmer/Zotero/storage/CEBUDJT7/41059.html}
+}
+
+@online{fiedlerAcceleratingEquilibrationFirstPrinciples2022,
+ title = {Accelerating {{Equilibration}} in {{First-Principles Molecular Dynamics}} with {{Orbital-Free Density Functional Theory}}},
+ author = {Fiedler, Lenz and Moldabekov, Zhandos A. and Shao, Xuecheng and Jiang, Kaili and Dornheim, Tobias and Pavanello, Michele and Cangi, Attila},
+ date = {2022-09-02},
+ number = {arXiv:2206.03754},
+ eprint = {arXiv:2206.03754},
+ eprinttype = {arxiv},
+ doi = {10.48550/arXiv.2206.03754},
+ url = {http://arxiv.org/abs/2206.03754},
+ urldate = {2023-02-15},
+ abstract = {We introduce a practical hybrid approach that combines orbital-free density functional theory (DFT) with Kohn-Sham DFT for speeding up first-principles molecular dynamics simulations. Equilibrated ionic configurations are generated using orbital-free DFT for subsequent Kohn-Sham DFT molecular dynamics. This leads to a massive reduction of the simulation time without any sacrifice in accuracy. We assess this finding across systems of different sizes and temperature, up to the warm dense matter regime. To that end, we use the cosine distance between the time series of radial distribution functions representing the ionic configurations. Likewise, we show that the equilibrated ionic configurations from this hybrid approach significantly enhance the accuracy of machine-learning models that replace Kohn-Sham DFT. Our hybrid scheme enables systematic first-principles simulations of warm dense matter that are otherwise hampered by the large numbers of atoms and the prevalent high temperatures. Moreover, our finding provides an additional motivation for developing kinetic and noninteracting free energy functionals for orbital-free DFT.},
+ pubstate = {preprint},
+ keywords = {/unread},
+ file = {/Users/wasmer/Nextcloud/Zotero/Fiedler et al_2022_Accelerating Equilibration in First-Principles Molecular Dynamics with.pdf;/Users/wasmer/Zotero/storage/TA7XVJUP/2206.html}
}
@unpublished{fiedlerDeepDiveMachine2021,
@@ -2284,13 +2503,82 @@
date = {2021-10-03},
eprint = {2110.00997},
eprinttype = {arxiv},
- primaryclass = {cond-mat},
+ eprintclass = {cond-mat},
url = {http://arxiv.org/abs/2110.00997},
urldate = {2021-11-17},
abstract = {With the growth of computational resources, the scope of electronic structure simulations has increased greatly. Artificial intelligence and robust data analysis hold the promise to accelerate large-scale simulations and their analysis to hitherto unattainable scales. Machine learning is a rapidly growing field for the processing of such complex datasets. It has recently gained traction in the domain of electronic structure simulations, where density functional theory takes the prominent role of the most widely used electronic structure method. Thus, DFT calculations represent one of the largest loads on academic high-performance computing systems across the world. Accelerating these with machine learning can reduce the resources required and enables simulations of larger systems. Hence, the combination of density functional theory and machine learning has the potential to rapidly advance electronic structure applications such as in-silico materials discovery and the search for new chemical reaction pathways. We provide the theoretical background of both density functional theory and machine learning on a generally accessible level. This serves as the basis of our comprehensive review including research articles up to December 2020 in chemistry and materials science that employ machine-learning techniques. In our analysis, we categorize the body of research into main threads and extract impactful results. We conclude our review with an outlook on exciting research directions in terms of a citation analysis.},
- archiveprefix = {arXiv},
- keywords = {_tablet,Condensed Matter - Materials Science,DFT,ML,ML-DFT,ML-ESM,review},
- file = {/home/johannes/Nextcloud/Zotero/Fiedler et al_2021_A Deep Dive into Machine Learning Density Functional Theory for Materials.pdf;/home/johannes/Zotero/storage/2XW6IGEA/2110.html}
+ keywords = {_tablet,citation analysis,DFT,literature analysis,ML,ML-DFT,ML-ESM,review},
+ file = {/Users/wasmer/Nextcloud/Zotero/Fiedler et al_2021_A Deep Dive into Machine Learning Density Functional Theory for Materials.pdf;/Users/wasmer/Zotero/storage/2XW6IGEA/2110.html}
+}
+
+@online{fiedlerDeepDiveMachine2022,
+ title = {A {{Deep Dive}} into {{Machine Learning Density Functional Theory}} for {{Materials Science}} and {{Chemistry}}},
+ author = {Fiedler, Lenz and Shah, Karan and Bussmann, Michael and Cangi, Attila},
+ date = {2022-02-25},
+ number = {arXiv:2110.00997},
+ eprint = {arXiv:2110.00997},
+ eprinttype = {arxiv},
+ doi = {10.48550/arXiv.2110.00997},
+ url = {http://arxiv.org/abs/2110.00997},
+ urldate = {2023-02-15},
+ abstract = {With the growth of computational resources, the scope of electronic structure simulations has increased greatly. Artificial intelligence and robust data analysis hold the promise to accelerate large-scale simulations and their analysis to hitherto unattainable scales. Machine learning is a rapidly growing field for the processing of such complex datasets. It has recently gained traction in the domain of electronic structure simulations, where density functional theory takes the prominent role of the most widely used electronic structure method. Thus, DFT calculations represent one of the largest loads on academic high-performance computing systems across the world. Accelerating these with machine learning can reduce the resources required and enables simulations of larger systems. Hence, the combination of density functional theory and machine learning has the potential to rapidly advance electronic structure applications such as in-silico materials discovery and the search for new chemical reaction pathways. We provide the theoretical background of both density functional theory and machine learning on a generally accessible level. This serves as the basis of our comprehensive review including research articles up to December 2020 in chemistry and materials science that employ machine-learning techniques. In our analysis, we categorize the body of research into main threads and extract impactful results. We conclude our review with an outlook on exciting research directions in terms of a citation analysis.},
+ pubstate = {preprint},
+ keywords = {citation analysis,DFT,literature analysis,ML,ML-DFT,ML-ESM,review},
+ file = {/Users/wasmer/Nextcloud/Zotero/Fiedler et al_2022_A Deep Dive into Machine Learning Density Functional Theory for Materials.pdf;/Users/wasmer/Zotero/storage/NUWIJFTB/2110.html}
+}
+
+@article{fiedlerDeepDiveMachine2022a,
+ title = {Deep Dive into Machine Learning Density Functional Theory for Materials Science and Chemistry},
+ author = {Fiedler, L. and Shah, K. and Bussmann, M. and Cangi, A.},
+ date = {2022-04-05},
+ journaltitle = {Physical Review Materials},
+ shortjournal = {Phys. Rev. Mater.},
+ volume = {6},
+ number = {4},
+ pages = {040301},
+ publisher = {{American Physical Society}},
+ doi = {10.1103/PhysRevMaterials.6.040301},
+ url = {https://link.aps.org/doi/10.1103/PhysRevMaterials.6.040301},
+ urldate = {2023-03-09},
+ abstract = {With the growth of computational resources, the scope of electronic structure simulations has increased greatly. Artificial intelligence and robust data analysis hold the promise to accelerate large-scale simulations and their analysis to hitherto unattainable scales. Machine learning is a rapidly growing field for the processing of such complex data sets. It has recently gained traction in the domain of electronic structure simulations, where density functional theory (DFT) takes the prominent role of the most widely used electronic structure method. Thus, DFT calculations represent one of the largest loads on academic high-performance computing systems across the world. Accelerating these with machine learning can reduce the resources required and enables simulations of larger systems. Hence, the combination of DFT and machine learning has the potential to rapidly advance electronic structure applications such as in silico materials discovery and the search for new chemical reaction pathways. We provide the theoretical background of both DFT and machine learning on a generally accessible level. This serves as the basis of our comprehensive review, including research articles up to December 2020 in chemistry and materials science that employ machine-learning techniques. In our analysis, we categorize the body of research into main threads and extract impactful results. We conclude our review with an outlook on exciting research directions in terms of a citation analysis.},
+ keywords = {/unread,citation analysis,DFT,literature analysis,ML,ML-DFT,ML-ESM,review},
+ file = {/Users/wasmer/Nextcloud/Zotero/Fiedler et al_2022_Deep dive into machine learning density functional theory for materials science.pdf;/Users/wasmer/Zotero/storage/62FHUUPB/PhysRevMaterials.6.html}
+}
+
+@online{fiedlerPredictingElectronicStructures2022,
+ title = {Predicting Electronic Structures at Any Length Scale with Machine Learning},
+ author = {Fiedler, Lenz and Modine, Normand A. and Schmerler, Steve and Vogel, Dayton J. and Popoola, Gabriel A. and Thompson, Aidan P. and Rajamanickam, Sivasankaran and Cangi, Attila},
+ date = {2022-12-08},
+ number = {arXiv:2210.11343},
+ eprint = {arXiv:2210.11343},
+ eprinttype = {arxiv},
+ doi = {10.48550/arXiv.2210.11343},
+ url = {http://arxiv.org/abs/2210.11343},
+ urldate = {2023-02-15},
+ abstract = {The properties of electrons in matter are of fundamental importance. They give rise to virtually all molecular and material properties and determine the physics at play in objects ranging from semiconductor devices to the interior of giant gas planets. Modeling and simulation of such diverse applications rely primarily on density functional theory (DFT), which has become the principal method for predicting the electronic structure of matter. While DFT calculations have proven to be very useful to the point of being recognized with a Nobel prize in 1998, their computational scaling limits them to small systems. We have developed a machine learning framework for predicting the electronic structure on any length scale. It shows up to three orders of magnitude speedup on systems where DFT is tractable and, more importantly, enables predictions on scales where DFT calculations are infeasible. Our work demonstrates how machine learning circumvents a long-standing computational bottleneck and advances science to frontiers intractable with any current solutions. This unprecedented modeling capability opens up an inexhaustible range of applications in astrophysics, novel materials discovery, and energy solutions for a sustainable future.},
+ pubstate = {preprint},
+ file = {/Users/wasmer/Nextcloud/Zotero/Fiedler et al_2022_Predicting electronic structures at any length scale with machine learning.pdf;/Users/wasmer/Zotero/storage/9AYDDQ8T/2210.html}
+}
+
+@article{fiedlerTrainingfreeHyperparameterOptimization2022,
+ title = {Training-Free Hyperparameter Optimization of Neural Networks for Electronic Structures in Matter},
+ author = {Fiedler, Lenz and Hoffmann, Nils and Mohammed, Parvez and Popoola, Gabriel A. and Yovell, Tamar and Oles, Vladyslav and Ellis, J. Austin and Rajamanickam, Siva and Cangi, Attila},
+ date = {2022-12-01},
+ journaltitle = {Machine Learning: Science and Technology},
+ shortjournal = {Mach. Learn.: Sci. Technol.},
+ volume = {3},
+ number = {4},
+ eprint = {2202.09186},
+ eprinttype = {arxiv},
+ eprintclass = {cond-mat},
+ pages = {045008},
+ issn = {2632-2153},
+ doi = {10.1088/2632-2153/ac9956},
+ url = {http://arxiv.org/abs/2202.09186},
+ urldate = {2023-02-15},
+ abstract = {A myriad of phenomena in materials science and chemistry rely on quantum-level simulations of the electronic structure in matter. While moving to larger length and time scales has been a pressing issue for decades, such large-scale electronic structure calculations are still challenging despite modern software approaches and advances in high-performance computing. The silver lining in this regard is the use of machine learning to accelerate electronic structure calculations -- this line of research has recently gained growing attention. The grand challenge therein is finding a suitable machine-learning model during a process called hyperparameter optimization. This, however, causes a massive computational overhead in addition to that of data generation. We accelerate the construction of neural network models by roughly two orders of magnitude by circumventing excessive training during the hyperparameter optimization phase. We demonstrate our workflow for Kohn-Sham density functional theory, the most popular computational method in materials science and chemistry.},
+ keywords = {_tablet,/unread},
+ file = {/Users/wasmer/Nextcloud/Zotero/Fiedler et al_2022_Training-free hyperparameter optimization of neural networks for electronic.pdf;/Users/wasmer/Zotero/storage/6BKNM2VX/2202.html}
}
@article{flores-livasPredictionHotSuperconductivity2019,
@@ -2307,7 +2595,23 @@
abstract = {A proposed hydrogen-rich solid would superconduct above the boiling point of water—though the material would need to be subjected to a colossal pressure.},
langid = {english},
keywords = {applications of DFT,DFT,master-thesis,superconductor},
- file = {/home/johannes/Nextcloud/Zotero/Flores-Livas_Arita_2019_A Prediction for “Hot†Superconductivity.pdf;/home/johannes/Zotero/storage/IEKIVRSH/96.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Flores-Livas_Arita_2019_A Prediction for “Hot†Superconductivity.pdf;/Users/wasmer/Zotero/storage/IEKIVRSH/96.html}
+}
+
+@online{focassioLinearJacobiLegendreExpansion2023,
+ title = {Linear {{Jacobi-Legendre}} Expansion of the Charge Density for Machine Learning-Accelerated Electronic Structure Calculations},
+ author = {Focassio, Bruno and Domina, Michelangelo and Patil, Urvesh and Fazzio, Adalberto and Sanvito, Stefano},
+ date = {2023-01-31},
+ number = {arXiv:2301.13550},
+ eprint = {arXiv:2301.13550},
+ eprinttype = {arxiv},
+ doi = {10.48550/arXiv.2301.13550},
+ url = {http://arxiv.org/abs/2301.13550},
+ urldate = {2023-02-23},
+ abstract = {As the go-to method to solve the electronic structure problem, Kohn-Sham density functional theory (KS-DFT) can be used to obtain the ground-state charge density, total energy, and several other key materials' properties. Unfortunately, the solution of the Kohn-Sham equations is found iteratively. This is a numerically intensive task, limiting the possible size and complexity of the systems to be treated. Machine-learning (ML) models for the charge density can then be used as surrogates to generate the converged charge density and reduce the computational cost of solving the electronic structure problem. We derive a powerful grid-centred structural representation based on the Jacobi and Legendre polynomials that, combined with a linear regression built on a data-efficient workflow, can accurately learn the charge density. Then, we design a machine-learning pipeline that can return energy and forces at the quality of a converged DFT calculation but at a fraction of the computational cost. This can be used as a tool for the fast scanning of the energy landscape and as a starting point to the DFT self-consistent cycle, in both cases maintaining a low computational cost.},
+ pubstate = {preprint},
+ keywords = {_tablet,Condensed Matter - Materials Science,prediction of electron density},
+ file = {/Users/wasmer/Nextcloud/Zotero/Focassio et al_2023_Linear Jacobi-Legendre expansion of the charge density for machine.pdf;/Users/wasmer/Zotero/storage/HPSZ89R2/2301.html}
}
@report{foulkesTopologyEntanglementStrong2020,
@@ -2326,7 +2630,7 @@
isbn = {9783958064669},
langid = {english},
keywords = {_tablet,ACE,cusps,FermiNet,ML-ESM,ML-QMBP,prediction of wavefunction,QMC,Slater-Jastrow,VMC},
- file = {/home/johannes/Nextcloud/Zotero/Foulkes_Drautz_2020_Topology, Entanglement, and Strong Correlations.pdf;/home/johannes/Zotero/storage/WLIE37SZ/884084.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Foulkes_Drautz_2020_Topology, Entanglement, and Strong Correlations.pdf;/Users/wasmer/Zotero/storage/WLIE37SZ/884084.html}
}
@unpublished{frankDetectInteractionsThat2021,
@@ -2337,15 +2641,14 @@
number = {arXiv:2106.02549},
eprint = {2106.02549},
eprinttype = {arxiv},
- primaryclass = {physics},
+ eprintclass = {physics},
publisher = {{arXiv}},
doi = {10.48550/arXiv.2106.02549},
url = {http://arxiv.org/abs/2106.02549},
urldate = {2022-05-18},
abstract = {Attention mechanisms are developing into a viable alternative to convolutional layers as elementary building block of NNs. Their main advantage is that they are not restricted to capture local dependencies in the input, but can draw arbitrary connections. This unprecedented capability coincides with the long-standing problem of modeling global atomic interactions in molecular force fields and other many-body problems. In its original formulation, however, attention is not applicable to the continuous domains in which the atoms live. For this purpose we propose a variant to describe geometric relations for arbitrary atomic configurations in Euclidean space that also respects all relevant physical symmetries. We furthermore demonstrate, how the successive application of our learned attention matrices effectively translates the molecular geometry into a set of individual atomic contributions on-the-fly.},
- archiveprefix = {arXiv},
keywords = {attention,invariance,ML},
- file = {/home/johannes/Nextcloud/Zotero/Frank_Chmiela_2021_Detect the Interactions that Matter in Matter.pdf;/home/johannes/Zotero/storage/7QC4UBJN/2106.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Frank_Chmiela_2021_Detect the Interactions that Matter in Matter.pdf;/Users/wasmer/Zotero/storage/7QC4UBJN/2106.html}
}
@article{frauxChemiscopeInteractiveStructureproperty2020,
@@ -2363,8 +2666,8 @@
urldate = {2022-10-03},
abstract = {Fraux et al., (2020). Chemiscope: interactive structure-property explorer for materials and molecules. Journal of Open Source Software, 5(51), 2117, https://doi.org/10.21105/joss.02117},
langid = {english},
- keywords = {data exploration,Database,ML,molecules,sketchmap,solids,unsupervised learning,visualization},
- file = {/home/johannes/Nextcloud/Zotero/Fraux et al_2020_Chemiscope.pdf;/home/johannes/Zotero/storage/TCQI9XE2/joss.html}
+ keywords = {_tablet,data exploration,Database,ML,molecules,sketchmap,solids,unsupervised learning,visualization},
+ file = {/Users/wasmer/Nextcloud/Zotero/Fraux et al_2020_Chemiscope.pdf;/Users/wasmer/Zotero/storage/TCQI9XE2/joss.html}
}
@article{freyMachineLearningEnabledDesign2020,
@@ -2383,7 +2686,7 @@
urldate = {2021-05-20},
abstract = {Engineered point defects in two-dimensional (2D) materials offer an attractive platform for solid-state devices that exploit tailored optoelectronic, quantum emission, and resistive properties. Naturally occurring defects are also unavoidably important contributors to material properties and performance. The immense variety and complexity of possible defects make it challenging to experimentally control, probe, or understand atomic-scale defect-property relationships. Here, we develop an approach based on deep transfer learning, machine learning, and first-principles calculations to rapidly predict key properties of point defects in 2D materials. We use physics-informed featurization to generate a minimal description of defect structures and present a general picture of defects across materials systems. We identify over one hundred promising, unexplored dopant defect structures in layered metal chalcogenides, hexagonal nitrides, and metal halides. These defects are prime candidates for quantum emission, resistive switching, and neuromorphic computing.},
keywords = {2D material,classification,deep transfer learning,defects,DFT,materials discovery,materials for neuromorphic computing,materials for quantum computing,NN,random forest,regression,transfer learning},
- file = {/home/johannes/Nextcloud/Zotero/Frey et al_2020_Machine Learning-Enabled Design of Point Defects in 2D Materials for Quantum.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Frey et al_2020_Machine Learning-Enabled Design of Point Defects in 2D Materials for Quantum.pdf}
}
@article{freyNeuralScalingDeep2022,
@@ -2396,7 +2699,7 @@
abstract = {Massive scale, both in terms of data availability and computation, enables significant breakthroughs in key application areas of deep learning such as natural language processing (NLP) and computer vision. There is emerging evidence that scale may be a key ingredient in scientific deep learning, but the importance of physical priors in scientific domains makes the strategies and benefits of scaling uncertain. Here, we investigate neural scaling behavior in large chemical models by varying model and dataset sizes over many orders of magnitude, studying models with over one billion parameters, pre-trained on datasets of up to ten million datapoints. We consider large language models for generative chemistry and graph neural networks for machine-learned interatomic potentials. To enable large-scale scientific deep learning studies under resource constraints, we develop the Training Performance Estimation (TPE) framework to reduce the costs of scalable hyperparameter optimization by up to 90\%. Using this framework, we discover empirical neural scaling relations for deep chemical models and investigate the interplay between physical priors and scale. Potential applications of large, pre-trained models for "prompt engineering" and unsupervised representation learning of molecules are shown.},
langid = {english},
keywords = {Allegro,ChemGPT,hyperparameters optimization,large language models,large models,MLP,models comparison,original publication,PAiNN,scaling,SchNet,SpookyNet,TODO,Training Performance Estimation,with-code},
- file = {/home/johannes/Nextcloud/Zotero/Frey et al_2022_Neural Scaling of Deep Chemical Models.pdf;/home/johannes/Zotero/storage/RLZJQP57/627bddd544bdd532395fb4b5.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Frey et al_2022_Neural Scaling of Deep Chemical Models.pdf;/Users/wasmer/Zotero/storage/RLZJQP57/627bddd544bdd532395fb4b5.html}
}
@article{freysoldtFirstprinciplesCalculationsPoint2014,
@@ -2410,7 +2713,7 @@
pages = {253--305},
doi = {10.1103/RevModPhys.86.253},
keywords = {basics,defects,DFT,impurity embedding,MPI Eisenforschung,review},
- file = {/home/johannes/Zotero/storage/26HPBYJN/RevModPhys.86.html}
+ file = {/Users/wasmer/Zotero/storage/26HPBYJN/RevModPhys.86.html}
}
@article{friederichMachinelearnedPotentialsNextgeneration2021,
@@ -2431,7 +2734,7 @@
issue = {6},
langid = {english},
keywords = {descriptors,MD,ML,MLP,review,review-of-MLP},
- file = {/home/johannes/Nextcloud/Zotero/Friederich et al_2021_Machine-learned potentials for next-generation matter simulations.pdf;/home/johannes/Zotero/storage/KBK5DLLS/s41563-020-0777-6.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Friederich et al_2021_Machine-learned potentials for next-generation matter simulations.pdf;/Users/wasmer/Zotero/storage/KBK5DLLS/s41563-020-0777-6.html}
}
@article{frolovQuantumComputingReproducibility2021,
@@ -2452,26 +2755,24 @@
langid = {english},
keywords = {Majorana,Peer review,quantum computing,rec-by-ghosh,reproducibility crisis,skeptics},
annotation = {Bandiera\_abtest: a Cg\_type: Comment Subject\_term: Quantum physics, Publishing, Peer review},
- file = {/home/johannes/Nextcloud/Zotero/Frolov_2021_Quantum computing’s reproducibility crisis.pdf;/home/johannes/Zotero/storage/CLEGVGB5/d41586-021-00954-8.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Frolov_2021_Quantum computing’s reproducibility crisis.pdf;/Users/wasmer/Zotero/storage/CLEGVGB5/d41586-021-00954-8.html}
}
-@misc{fuchsSETransformers3D2020,
+@online{fuchsSETransformers3D2020,
title = {{{SE}}(3)-{{Transformers}}: {{3D Roto-Translation Equivariant Attention Networks}}},
shorttitle = {{{SE}}(3)-{{Transformers}}},
author = {Fuchs, Fabian B. and Worrall, Daniel E. and Fischer, Volker and Welling, Max},
date = {2020-11-24},
number = {arXiv:2006.10503},
- eprint = {2006.10503},
+ eprint = {arXiv:2006.10503},
eprinttype = {arxiv},
- primaryclass = {cs, stat},
- publisher = {{arXiv}},
doi = {10.48550/arXiv.2006.10503},
url = {http://arxiv.org/abs/2006.10503},
urldate = {2022-10-03},
abstract = {We introduce the SE(3)-Transformer, a variant of the self-attention module for 3D point clouds and graphs, which is equivariant under continuous 3D roto-translations. Equivariance is important to ensure stable and predictable performance in the presence of nuisance transformations of the data input. A positive corollary of equivariance is increased weight-tying within the model. The SE(3)-Transformer leverages the benefits of self-attention to operate on large point clouds and graphs with varying number of points, while guaranteeing SE(3)-equivariance for robustness. We evaluate our model on a toy N-body particle simulation dataset, showcasing the robustness of the predictions under rotations of the input. We further achieve competitive performance on two real-world datasets, ScanObjectNN and QM9. In all cases, our model outperforms a strong, non-equivariant attention baseline and an equivariant model without attention.},
- archiveprefix = {arXiv},
+ pubstate = {preprint},
keywords = {equivariant,GCN,GNN,library,ML,QM9,SchNet,SE(3),self-attention,transformer,with-code},
- file = {/home/johannes/Nextcloud/Zotero/Fuchs et al_2020_SE(3)-Transformers.pdf;/home/johannes/Zotero/storage/UMVV286P/2006.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Fuchs et al_2020_SE(3)-Transformers.pdf;/Users/wasmer/Zotero/storage/UMVV286P/2006.html}
}
@article{fungPhysicallyInformedMachine2022,
@@ -2489,7 +2790,7 @@
url = {https://doi.org/10.1021/acs.chemmater.1c04252},
urldate = {2022-07-10},
abstract = {The electronic structure of a material, such as its density of states (DOS), provides key insights into its physical and functional properties and serves as a valuable source of high-quality features for many materials screening and discovery workflows. However, the computational cost of calculating the DOS, most commonly with density functional theory (DFT), becomes prohibitive for meeting high-fidelity or high-throughput requirements, necessitating a cheaper but sufficiently accurate surrogate. To fulfill this demand, we develop a general machine learning method based on graph neural networks for predicting the DOS purely from atomic positions, six orders of magnitude faster than DFT. This approach can effectively use large materials databases and be applied generally across the entire periodic table to materials classes of arbitrary compositional and structural diversity. We furthermore devise a highly adaptable scheme for physically informed learning which encourages the DOS prediction to favor physically reasonable solutions defined by any set of desired constraints. This functionality provides a means for ensuring that the predicted DOS is reliable enough to be used as an input to downstream materials screening workflows to predict more complex functional properties, which rely on accurate physical features.},
- file = {/home/johannes/Nextcloud/Zotero/Fung et al_2022_Physically Informed Machine Learning Prediction of Electronic Density of States.pdf;/home/johannes/Zotero/storage/MFQH6849/acs.chemmater.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Fung et al_2022_Physically Informed Machine Learning Prediction of Electronic Density of States.pdf;/Users/wasmer/Zotero/storage/MFQH6849/acs.chemmater.html}
}
@online{galkinGraphML20222021,
@@ -2502,7 +2803,7 @@
abstract = {Hot trends and major advancements},
langid = {english},
organization = {{Medium}},
- file = {/home/johannes/Zotero/storage/8ESSCXA2/graph-ml-in-2022-where-are-we-now-f7f8242599e0.html}
+ file = {/Users/wasmer/Zotero/storage/8ESSCXA2/graph-ml-in-2022-where-are-we-now-f7f8242599e0.html}
}
@book{gammaDesignPatternsElements1995,
@@ -2518,22 +2819,36 @@
keywords = {OO,Reusability,software engineering,Software patterns}
}
-@misc{gardnerSyntheticDataEnable2022,
+@online{gardnerSyntheticDataEnable2022,
title = {Synthetic Data Enable Experiments in Atomistic Machine Learning},
author = {Gardner, John L. A. and Beaulieu, Zoé Faure and Deringer, Volker L.},
date = {2022-11-29},
number = {arXiv:2211.16443},
- eprint = {2211.16443},
+ eprint = {arXiv:2211.16443},
eprinttype = {arxiv},
- primaryclass = {physics},
- publisher = {{arXiv}},
doi = {10.48550/arXiv.2211.16443},
url = {http://arxiv.org/abs/2211.16443},
urldate = {2022-12-29},
abstract = {Machine-learning models are increasingly used to predict properties of atoms in chemical systems. There have been major advances in developing descriptors and regression frameworks for this task, typically starting from (relatively) small sets of quantum-mechanical reference data. Larger datasets of this kind are becoming available, but remain expensive to generate. Here we demonstrate the use of a large dataset that we have "synthetically" labelled with per-atom energies from an existing ML potential model. The cheapness of this process, compared to the quantum-mechanical ground truth, allows us to generate millions of datapoints, in turn enabling rapid experimentation with atomistic ML models from the small- to the large-data regime. This approach allows us here to compare regression frameworks in depth, and to explore visualisation based on learned representations. We also show that learning synthetic data labels can be a useful pre-training task for subsequent fine-tuning on small datasets. In the future, we expect that our open-sourced dataset, and similar ones, will be useful in rapidly exploring deep-learning models in the limit of abundant chemical data.},
- archiveprefix = {arXiv},
- keywords = {data augmentation,GAP,GPR,MD,ML,MLP,NN,prediction of potential energy,small data,SOAP,synthetic data},
- file = {/home/johannes/Nextcloud/Zotero/Gardner et al_2022_Synthetic data enable experiments in atomistic machine learning.pdf;/home/johannes/Zotero/storage/N3NP679J/2211.html}
+ pubstate = {preprint},
+ keywords = {_tablet,data augmentation,GAP,GPR,MD,ML,MLP,NN,prediction of potential energy,small data,SOAP,synthetic data},
+ file = {/Users/wasmer/Nextcloud/Zotero/Gardner et al_2022_Synthetic data enable experiments in atomistic machine learning.pdf;/Users/wasmer/Zotero/storage/N3NP679J/2211.html}
+}
+
+@online{gardnerSyntheticDataEnable2022a,
+ title = {Synthetic Data Enable Experiments in Atomistic Machine Learning},
+ author = {Gardner, John L. A. and Beaulieu, Zoé Faure and Deringer, Volker L.},
+ date = {2022-11-29},
+ number = {arXiv:2211.16443},
+ eprint = {arXiv:2211.16443},
+ eprinttype = {arxiv},
+ doi = {10.48550/arXiv.2211.16443},
+ url = {http://arxiv.org/abs/2211.16443},
+ urldate = {2023-03-01},
+ abstract = {Machine-learning models are increasingly used to predict properties of atoms in chemical systems. There have been major advances in developing descriptors and regression frameworks for this task, typically starting from (relatively) small sets of quantum-mechanical reference data. Larger datasets of this kind are becoming available, but remain expensive to generate. Here we demonstrate the use of a large dataset that we have "synthetically" labelled with per-atom energies from an existing ML potential model. The cheapness of this process, compared to the quantum-mechanical ground truth, allows us to generate millions of datapoints, in turn enabling rapid experimentation with atomistic ML models from the small- to the large-data regime. This approach allows us here to compare regression frameworks in depth, and to explore visualisation based on learned representations. We also show that learning synthetic data labels can be a useful pre-training task for subsequent fine-tuning on small datasets. In the future, we expect that our open-sourced dataset, and similar ones, will be useful in rapidly exploring deep-learning models in the limit of abundant chemical data.},
+ pubstate = {preprint},
+ keywords = {/unread,carbon,data augmentation,Database,disordered,DKL,GAP,GPR,KPCovR,LAMMPS,MD,MLP,NN,SOAP,Supervised learning,synthetic data,UMAP,unsupervised learning},
+ file = {/Users/wasmer/Nextcloud/Zotero/Gardner et al_2022_Synthetic data enable experiments in atomistic machine learning2.pdf;/Users/wasmer/Zotero/storage/99FPUBGW/2211.html}
}
@article{garridotorresLowScalingAlgorithmNudged2019,
@@ -2551,80 +2866,72 @@
urldate = {2021-08-21},
abstract = {We present the incorporation of a surrogate Gaussian process regression (GPR) atomistic model to greatly accelerate the rate of convergence of classical nudged elastic band (NEB) calculations. In our surrogate model approach, the cost of converging the elastic band no longer scales with the number of moving images on the path. This provides a far more efficient and robust transition state search. In contrast to a conventional NEB calculation, the algorithm presented here eliminates any need for manipulating the number of images to obtain a converged result. This is achieved by inventing a new convergence criteria that exploits the probabilistic nature of the GPR to use uncertainty estimates of all images in combination with the force in the saddle point in the target model potential. Our method is an order of magnitude faster in terms of function evaluations than the conventional NEB method with no accuracy loss for the converged energy barrier values.},
keywords = {DFT,GPR,ML,models,NEB,rec-by-ruess,surrogate model},
- file = {/home/johannes/Nextcloud/Zotero/Garrido Torres et al_2019_Low-Scaling Algorithm for Nudged Elastic Band Calculations Using a Surrogate.pdf;/home/johannes/Zotero/storage/TFSWCWBH/Garrido Torres et al. - 2019 - Low-Scaling Algorithm for Nudged Elastic Band Calc.pdf;/home/johannes/Zotero/storage/DWT7X58R/PhysRevLett.122.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Garrido Torres et al_2019_Low-Scaling Algorithm for Nudged Elastic Band Calculations Using a Surrogate.pdf;/Users/wasmer/Zotero/storage/TFSWCWBH/Garrido Torres et al. - 2019 - Low-Scaling Algorithm for Nudged Elastic Band Calc.pdf;/Users/wasmer/Zotero/storage/DWT7X58R/PhysRevLett.122.html}
}
-@misc{gasteigerDirectionalMessagePassing2022,
+@online{gasteigerDirectionalMessagePassing2022,
title = {Directional {{Message Passing}} for {{Molecular Graphs}}},
author = {Gasteiger, Johannes and Groß, Janek and Günnemann, Stephan},
date = {2022-04-05},
number = {arXiv:2003.03123},
- eprint = {2003.03123},
+ eprint = {arXiv:2003.03123},
eprinttype = {arxiv},
- primaryclass = {physics, stat},
- publisher = {{arXiv}},
doi = {10.48550/arXiv.2003.03123},
url = {http://arxiv.org/abs/2003.03123},
urldate = {2022-10-03},
abstract = {Graph neural networks have recently achieved great successes in predicting quantum mechanical properties of molecules. These models represent a molecule as a graph using only the distance between atoms (nodes). They do not, however, consider the spatial direction from one atom to another, despite directional information playing a central role in empirical potentials for molecules, e.g. in angular potentials. To alleviate this limitation we propose directional message passing, in which we embed the messages passed between atoms instead of the atoms themselves. Each message is associated with a direction in coordinate space. These directional message embeddings are rotationally equivariant since the associated directions rotate with the molecule. We propose a message passing scheme analogous to belief propagation, which uses the directional information by transforming messages based on the angle between them. Additionally, we use spherical Bessel functions and spherical harmonics to construct theoretically well-founded, orthogonal representations that achieve better performance than the currently prevalent Gaussian radial basis representations while using fewer than 1/4 of the parameters. We leverage these innovations to construct the directional message passing neural network (DimeNet). DimeNet outperforms previous GNNs on average by 76\% on MD17 and by 31\% on QM9. Our implementation is available online.},
- archiveprefix = {arXiv},
+ pubstate = {preprint},
keywords = {DimeNet,GNN,MD,ML,MLP,molecules,MPNN,original publication,with-code},
- file = {/home/johannes/Nextcloud/Zotero/Gasteiger et al_2022_Directional Message Passing for Molecular Graphs.pdf;/home/johannes/Zotero/storage/G7KWBFCS/2003.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Gasteiger et al_2022_Directional Message Passing for Molecular Graphs.pdf;/Users/wasmer/Zotero/storage/G7KWBFCS/2003.html}
}
-@misc{gasteigerDirectionalMessagePassing2022a,
+@online{gasteigerDirectionalMessagePassing2022a,
title = {Directional {{Message Passing}} on {{Molecular Graphs}} via {{Synthetic Coordinates}}},
author = {Gasteiger, Johannes and Yeshwanth, Chandan and Günnemann, Stephan},
date = {2022-04-05},
number = {arXiv:2111.04718},
- eprint = {2111.04718},
+ eprint = {arXiv:2111.04718},
eprinttype = {arxiv},
- primaryclass = {physics, q-bio},
- publisher = {{arXiv}},
doi = {10.48550/arXiv.2111.04718},
url = {http://arxiv.org/abs/2111.04718},
urldate = {2022-10-03},
abstract = {Graph neural networks that leverage coordinates via directional message passing have recently set the state of the art on multiple molecular property prediction tasks. However, they rely on atom position information that is often unavailable, and obtaining it is usually prohibitively expensive or even impossible. In this paper we propose synthetic coordinates that enable the use of advanced GNNs without requiring the true molecular configuration. We propose two distances as synthetic coordinates: Distance bounds that specify the rough range of molecular configurations, and graph-based distances using a symmetric variant of personalized PageRank. To leverage both distance and angular information we propose a method of transforming normal graph neural networks into directional MPNNs. We show that with this transformation we can reduce the error of a normal graph neural network by 55\% on the ZINC benchmark. We furthermore set the state of the art on ZINC and coordinate-free QM9 by incorporating synthetic coordinates in the SMP and DimeNet++ models. Our implementation is available online.},
- archiveprefix = {arXiv},
+ pubstate = {preprint},
keywords = {DimeNet,DimeNet++,GNN,MD,MLP,molecules,MPNN,open-review},
- file = {/home/johannes/Nextcloud/Zotero/Gasteiger et al_2022_Directional Message Passing on Molecular Graphs via Synthetic Coordinates.pdf;/home/johannes/Zotero/storage/FEPN4JW4/2111.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Gasteiger et al_2022_Directional Message Passing on Molecular Graphs via Synthetic Coordinates.pdf;/Users/wasmer/Zotero/storage/FEPN4JW4/2111.html}
}
-@misc{gasteigerFastUncertaintyAwareDirectional2022,
+@online{gasteigerFastUncertaintyAwareDirectional2022,
title = {Fast and {{Uncertainty-Aware Directional Message Passing}} for {{Non-Equilibrium Molecules}}},
author = {Gasteiger, Johannes and Giri, Shankari and Margraf, Johannes T. and Günnemann, Stephan},
date = {2022-04-05},
number = {arXiv:2011.14115},
- eprint = {2011.14115},
+ eprint = {arXiv:2011.14115},
eprinttype = {arxiv},
- primaryclass = {physics},
- publisher = {{arXiv}},
doi = {10.48550/arXiv.2011.14115},
url = {http://arxiv.org/abs/2011.14115},
urldate = {2022-10-03},
abstract = {Many important tasks in chemistry revolve around molecules during reactions. This requires predictions far from the equilibrium, while most recent work in machine learning for molecules has been focused on equilibrium or near-equilibrium states. In this paper we aim to extend this scope in three ways. First, we propose the DimeNet++ model, which is 8x faster and 10\% more accurate than the original DimeNet on the QM9 benchmark of equilibrium molecules. Second, we validate DimeNet++ on highly reactive molecules by developing the challenging COLL dataset, which contains distorted configurations of small molecules during collisions. Finally, we investigate ensembling and mean-variance estimation for uncertainty quantification with the goal of accelerating the exploration of the vast space of non-equilibrium structures. Our DimeNet++ implementation as well as the COLL dataset are available online.},
- archiveprefix = {arXiv},
+ pubstate = {preprint},
keywords = {DimeNet,DimeNet++,GNN,MD,ML,MLP,molecules,MPNN,original publication,with-code},
- file = {/home/johannes/Nextcloud/Zotero/Gasteiger et al_2022_Fast and Uncertainty-Aware Directional Message Passing for Non-Equilibrium.pdf;/home/johannes/Zotero/storage/BVEXST79/2011.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Gasteiger et al_2022_Fast and Uncertainty-Aware Directional Message Passing for Non-Equilibrium.pdf;/Users/wasmer/Zotero/storage/BVEXST79/2011.html}
}
-@misc{gasteigerGemNetUniversalDirectional2022,
+@online{gasteigerGemNetUniversalDirectional2022,
title = {{{GemNet}}: {{Universal Directional Graph Neural Networks}} for {{Molecules}}},
shorttitle = {{{GemNet}}},
author = {Gasteiger, Johannes and Becker, Florian and Günnemann, Stephan},
date = {2022-04-05},
number = {arXiv:2106.08903},
- eprint = {2106.08903},
+ eprint = {arXiv:2106.08903},
eprinttype = {arxiv},
- primaryclass = {physics, stat},
- publisher = {{arXiv}},
doi = {10.48550/arXiv.2106.08903},
url = {http://arxiv.org/abs/2106.08903},
urldate = {2022-10-03},
abstract = {Effectively predicting molecular interactions has the potential to accelerate molecular dynamics by multiple orders of magnitude and thus revolutionize chemical simulations. Graph neural networks (GNNs) have recently shown great successes for this task, overtaking classical methods based on fixed molecular kernels. However, they still appear very limited from a theoretical perspective, since regular GNNs cannot distinguish certain types of graphs. In this work we close this gap between theory and practice. We show that GNNs with directed edge embeddings and two-hop message passing are indeed universal approximators for predictions that are invariant to translation, and equivariant to permutation and rotation. We then leverage these insights and multiple structural improvements to propose the geometric message passing neural network (GemNet). We demonstrate the benefits of the proposed changes in multiple ablation studies. GemNet outperforms previous models on the COLL, MD17, and OC20 datasets by 34\%, 41\%, and 20\%, respectively, and performs especially well on the most challenging molecules. Our implementation is available online.},
- archiveprefix = {arXiv},
+ pubstate = {preprint},
keywords = {DimeNet,DimeNet++,GemNet,MD,ML,MLP,molecules,MPNN,with-code},
- file = {/home/johannes/Nextcloud/Zotero/Gasteiger et al_2022_GemNet.pdf;/home/johannes/Zotero/storage/FE5R77B9/2106.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Gasteiger et al_2022_GemNet.pdf;/Users/wasmer/Zotero/storage/FE5R77B9/2106.html}
}
@article{gebauerInverseDesign3d2022,
@@ -2645,7 +2952,7 @@
issue = {1},
langid = {english},
keywords = {G-SchNet,generative models,inverse design,materials discovery,ML,prediction from properties,prediction of structure,SchNet},
- file = {/home/johannes/Nextcloud/Zotero/Gebauer et al_2022_Inverse design of 3d molecular structures with conditional generative neural.pdf;/home/johannes/Zotero/storage/XHLV2UHD/s41467-022-28526-y.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Gebauer et al_2022_Inverse design of 3d molecular structures with conditional generative neural.pdf;/Users/wasmer/Zotero/storage/XHLV2UHD/s41467-022-28526-y.html}
}
@article{gedeonMachineLearningDerivative2021,
@@ -2664,32 +2971,29 @@
urldate = {2022-01-10},
abstract = {Machine learning is a powerful tool to design accurate, highly non-local, exchange-correlation functionals for density functional theory. So far, most of those machine learned functionals are trained for systems with an integer number of particles. As such, they are unable to reproduce some crucial and fundamental aspects, such as the explicit dependency of the functionals on the particle number or the infamous derivative discontinuity at integer particle numbers. Here we propose a solution to these problems by training a neural network as the universal functional of density-functional theory that (a) depends explicitly on the number of particles with a piece-wise linearity between the integer numbers and (b) reproduces the derivative discontinuity of the exchange-correlation energy. This is achieved by using an ensemble formalism, a training set containing fractional densities, and an explicitly discontinuous formulation.},
langid = {english},
- file = {/home/johannes/Nextcloud/Zotero/Gedeon et al_2021_Machine learning the derivative discontinuity of density-functional theory.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Gedeon et al_2021_Machine learning the derivative discontinuity of density-functional theory.pdf}
}
-@misc{geigerE3nnEuclideanNeural2022,
+@online{geigerE3nnEuclideanNeural2022,
title = {E3nn: {{Euclidean Neural Networks}}},
shorttitle = {E3nn},
author = {Geiger, Mario and Smidt, Tess},
date = {2022-07-18},
number = {arXiv:2207.09453},
- eprint = {2207.09453},
+ eprint = {arXiv:2207.09453},
eprinttype = {arxiv},
- primaryclass = {cs},
- publisher = {{arXiv}},
doi = {10.48550/arXiv.2207.09453},
url = {http://arxiv.org/abs/2207.09453},
urldate = {2022-08-21},
abstract = {We present e3nn, a generalized framework for creating E(3) equivariant trainable functions, also known as Euclidean neural networks. e3nn naturally operates on geometry and geometric tensors that describe systems in 3D and transform predictably under a change of coordinate system. The core of e3nn are equivariant operations such as the TensorProduct class or the spherical harmonics functions that can be composed to create more complex modules such as convolutions and attention mechanisms. These core operations of e3nn can be used to efficiently articulate Tensor Field Networks, 3D Steerable CNNs, Clebsch-Gordan Networks, SE(3) Transformers and other E(3) equivariant networks.},
- archiveprefix = {arXiv},
+ pubstate = {preprint},
keywords = {_tablet,e3nn,EGNN,ENN,equivariant,library,ML-ESM,prediction of electron density,with-code},
- file = {/home/johannes/Nextcloud/Zotero/Geiger_Smidt_2022_e3nn.pdf;/home/johannes/Zotero/storage/SJW8392C/2207.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Geiger_Smidt_2022_e3nn.pdf;/Users/wasmer/Zotero/storage/SJW8392C/2207.html}
}
@article{genschComprehensiveDiscoveryPlatform2021,
title = {A {{Comprehensive Discovery Platform}} for {{Organophosphorus Ligands}} for {{Catalysis}}},
- author = {Gensch, Tobias and dos Passos Gomes, Gabriel and Friederich, Pascal and Peters, Ellyn and Gaudin, Theophile and Pollice, Robert and Jorner, Kjell and Nigam, AkshatKumar and Lindner D'Addario, Michael and Sigman, Matthew S. and Aspuru-Guzik, Alan},
- options = {useprefix=true},
+ author = {Gensch, Tobias and family=Passos Gomes, given=Gabriel, prefix=dos, useprefix=true and Friederich, Pascal and Peters, Ellyn and Gaudin, Theophile and Pollice, Robert and Jorner, Kjell and Nigam, AkshatKumar and Lindner D'Addario, Michael and Sigman, Matthew S. and Aspuru-Guzik, Alan},
date = {2021-04-27},
publisher = {{ChemRxiv}},
doi = {10.26434/chemrxiv.12996665.v1},
@@ -2698,26 +3002,24 @@
abstract = {The design of molecular catalysts typically involves reconciling multiple conflicting property requirements, largely relying on human intuition and local structural searches. However, the vast number of potential catalysts requires pruning of the candidate space by efficient property prediction with quantitative structure-property relationships. Data-driven workflows embedded in a library of potential catalysts can be used to build predictive models for catalyst performance and serve as a blueprint for novel catalyst designs. Herein we introduce kraken, a discovery platform covering monodentate organophosphorus(III) ligands providing comprehensive physicochemical descriptors based on representative conformer ensembles. Using quantum-mechanical methods, we calculated descriptors for 1,558 ligands, including commercially available examples, and trained machine learning models to predict properties of over 300,000 new ligands. We demonstrate the application of kraken to systematically explore the property space of organophosphorus ligands and how existing datasets in catalysis can be used to accelerate ligand selection during reaction optimization.},
langid = {english},
keywords = {chemistry,descriptors,kraken,materials database,ML,models,organic chemistry,visualization},
- file = {/home/johannes/Nextcloud/Zotero/Gensch et al_2021_A Comprehensive Discovery Platform for Organophosphorus Ligands for Catalysis.pdf;/home/johannes/Zotero/storage/ZSYETR3S/12996665.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Gensch et al_2021_A Comprehensive Discovery Platform for Organophosphorus Ligands for Catalysis.pdf;/Users/wasmer/Zotero/storage/ZSYETR3S/12996665.html}
}
-@misc{gerardGoldstandardSolutionsSchr2022,
+@online{gerardGoldstandardSolutionsSchr2022,
title = {Gold-Standard Solutions to the {{Schr}}\textbackslash "odinger Equation Using Deep Learning: {{How}} Much Physics Do We Need?},
shorttitle = {Gold-Standard Solutions to the {{Schr}}\textbackslash "odinger Equation Using Deep Learning},
author = {Gerard, Leon and Scherbela, Michael and Marquetand, Philipp and Grohs, Philipp},
date = {2022-05-31},
number = {arXiv:2205.09438},
- eprint = {2205.09438},
+ eprint = {arXiv:2205.09438},
eprinttype = {arxiv},
- primaryclass = {physics},
- publisher = {{arXiv}},
doi = {10.48550/arXiv.2205.09438},
url = {http://arxiv.org/abs/2205.09438},
urldate = {2022-08-16},
abstract = {Finding accurate solutions to the Schr\textbackslash "odinger equation is the key unsolved challenge of computational chemistry. Given its importance for the development of new chemical compounds, decades of research have been dedicated to this problem, but due to the large dimensionality even the best available methods do not yet reach the desired accuracy. Recently the combination of deep learning with Monte Carlo methods has emerged as a promising way to obtain highly accurate energies and moderate scaling of computational cost. In this paper we significantly contribute towards this goal by introducing a novel deep-learning architecture that achieves 40-70\% lower energy error at 8x lower computational cost compared to previous approaches. Using our method we establish a new benchmark by calculating the most accurate variational ground state energies ever published for a number of different atoms and molecules. We systematically break down and measure our improvements, focusing in particular on the effect of increasing physical prior knowledge. We surprisingly find that increasing the prior knowledge given to the architecture can actually decrease accuracy.},
- archiveprefix = {arXiv},
+ pubstate = {preprint},
keywords = {DNN,FermiNet,ML-ESM,ML-QMBP,molecules,PauliNet,prediction of wavefunction,QMC,VMC},
- file = {/home/johannes/Nextcloud/Zotero/Gerard et al_2022_Gold-standard solutions to the Schr-odinger equation using deep learning.pdf;/home/johannes/Zotero/storage/DWVRHXZW/2205.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Gerard et al_2022_Gold-standard solutions to the Schr-odinger equation using deep learning.pdf;/Users/wasmer/Zotero/storage/DWVRHXZW/2205.html}
}
@book{geronHandsonMachineLearning2019,
@@ -2751,7 +3053,7 @@
abstract = {Luca Ghiringhelli introduces an AI toolkit that can be used with materials databases to discover new materials, or new properties of known materials.},
langid = {english},
annotation = {Bandiera\_abtest: a Cg\_type: Nature Research Journals Primary\_atype: Research Highlights Subject\_term: Computational methods;Scientific data Subject\_term\_id: computational-methods;scientific-data},
- file = {/home/johannes/Nextcloud/Zotero/Ghiringhelli_2021_An AI-toolkit to develop and share research into new materials.pdf;/home/johannes/Zotero/storage/LTJNU3SG/s42254-021-00373-8.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Ghiringhelli_2021_An AI-toolkit to develop and share research into new materials.pdf;/Users/wasmer/Zotero/storage/LTJNU3SG/s42254-021-00373-8.html}
}
@article{ghiringhelliBigDataMaterials2015,
@@ -2770,7 +3072,7 @@
urldate = {2021-05-15},
abstract = {Statistical learning of materials properties or functions so far starts with a largely silent, nonchallenged step: the choice of the set of descriptive parameters (termed descriptor). However, when the scientific connection between the descriptor and the actuating mechanisms is unclear, the causality of the learned descriptor-property relation is uncertain. Thus, a trustful prediction of new promising materials, identification of anomalies, and scientific advancement are doubtful. We analyze this issue and define requirements for a suitable descriptor. For a classic example, the energy difference of zinc blende or wurtzite and rocksalt semiconductors, we demonstrate how a meaningful descriptor can be found systematically.},
keywords = {descriptors,ML},
- file = {/home/johannes/Nextcloud/Zotero/Ghiringhelli et al_2015_Big Data of Materials Science.pdf;/home/johannes/Zotero/storage/VMWGSVVB/PhysRevLett.114.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Ghiringhelli et al_2015_Big Data of Materials Science.pdf;/Users/wasmer/Zotero/storage/VMWGSVVB/PhysRevLett.114.html}
}
@article{ghiringhelliEfficientDataExchange2017,
@@ -2793,43 +3095,55 @@
langid = {english},
keywords = {materials database,materials informatics,materials metadata,NOMAD},
annotation = {Bandiera\_abtest: a Cc\_license\_type: cc\_by Cg\_type: Nature Research Journals Primary\_atype: Reviews Subject\_term: Condensed-matter physics;Theory and computation Subject\_term\_id: condensed-matter-physics;theory-and-computation},
- file = {/home/johannes/Nextcloud/Zotero/Ghiringhelli et al_2017_Towards efficient data exchange and sharing for big-data driven materials.pdf;/home/johannes/Zotero/storage/G3CTM9SN/s41524-017-0048-5.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Ghiringhelli et al_2017_Towards efficient data exchange and sharing for big-data driven materials.pdf;/Users/wasmer/Zotero/storage/G3CTM9SN/s41524-017-0048-5.html}
}
-@misc{ghoshClassicalQuantumMachine2022,
+@online{ghoshClassicalQuantumMachine2022,
title = {Classical and Quantum Machine Learning Applications in Spintronics},
author = {Ghosh, Kumar and Ghosh, Sumit},
date = {2022-07-26},
number = {arXiv:2207.12837},
- eprint = {2207.12837},
+ eprint = {arXiv:2207.12837},
eprinttype = {arxiv},
- primaryclass = {cond-mat, physics:physics, physics:quant-ph},
- publisher = {{arXiv}},
doi = {10.48550/arXiv.2207.12837},
url = {http://arxiv.org/abs/2207.12837},
urldate = {2022-10-28},
abstract = {In this article we demonstrate the applications of classical and quantum machine learning in quantum transport and spintronics. With the help of a two terminal device with magnetic impurity we show how machine learning algorithms can predict the highly non-linear nature of conductance as well as the non-equilibrium spin response function for any random magnetic configuration. We finally describe the applicability of quantum machine learning which has the capability to handle a significantly large configuration space. Our approach is also applicable for molecular systems. These outcomes are crucial in predicting the behaviour of large scale systems where a quantum mechanical calculation is computationally challenging and therefore would play a crucial role in designing nano devices.},
- archiveprefix = {arXiv},
- keywords = {ML,PGI-1/IAS-1,QML,QSVM,quantum computing,quantum transport,random forest,rec-by-ghosh,spin dynamics,Spintronics,SVM,tight binding,transport properties},
- file = {/home/johannes/Nextcloud/Zotero/Ghosh_Ghosh_2022_Classical and quantum machine learning applications in spintronics.pdf;/home/johannes/Zotero/storage/FEUD8XZQ/2207.html}
+ pubstate = {preprint},
+ keywords = {_tablet,ML,PGI-1/IAS-1,QML,QSVM,quantum computing,quantum transport,random forest,rec-by-ghosh,spin dynamics,Spintronics,SVM,tight binding,transport properties},
+ file = {/Users/wasmer/Nextcloud/Zotero/Ghosh_Ghosh_2022_Classical and quantum machine learning applications in spintronics.pdf;/Users/wasmer/Zotero/storage/FEUD8XZQ/2207.html}
}
-@misc{gilmerNeuralMessagePassing2017,
+@online{gilliganRulefreeWorkflowAutomated2023,
+ title = {A Rule-Free Workflow for the Automated Generation of Databases from Scientific Literature},
+ author = {Gilligan, Luke P. J. and Cobelli, Matteo and Taufour, Valentin and Sanvito, Stefano},
+ date = {2023-01-27},
+ number = {arXiv:2301.11689},
+ eprint = {arXiv:2301.11689},
+ eprinttype = {arxiv},
+ doi = {10.48550/arXiv.2301.11689},
+ url = {http://arxiv.org/abs/2301.11689},
+ urldate = {2023-02-23},
+ abstract = {In recent times, transformer networks have achieved state-of-the-art performance in a wide range of natural language processing tasks. Here we present a workflow based on the fine-tuning of BERT models for different downstream tasks, which results in the automated extraction of structured information from unstructured natural language in scientific literature. Contrary to other methods for the automated extraction of structured compound-property relations from similar sources, our workflow does not rely on the definition of intricate grammar rules. Hence, it can be adapted to a new task without requiring extensive implementation efforts and knowledge. We test the data extraction performance by automatically generating a database of compounds and their associated Curie temperatures. This is compared with a manually curated database and one obtained with the state-of-the-art rule-based method. Finally, in order to demonstrate that the automatically extracted database can be used in a material-design workflow, we employ it to construct a machine-learning model predicting the Curie temperature based on a compound's chemical composition. This is quantitatively tested and compared with the best model constructed on manually-extracted data.},
+ pubstate = {preprint},
+ keywords = {_tablet,/unread,Condensed Matter - Materials Science,Physics - Computational Physics,Physics - Data Analysis; Statistics and Probability},
+ file = {/Users/wasmer/Nextcloud/Zotero/Gilligan et al_2023_A rule-free workflow for the automated generation of databases from scientific.pdf;/Users/wasmer/Zotero/storage/W8WDMBDK/2301.html}
+}
+
+@online{gilmerNeuralMessagePassing2017,
title = {Neural {{Message Passing}} for {{Quantum Chemistry}}},
author = {Gilmer, Justin and Schoenholz, Samuel S. and Riley, Patrick F. and Vinyals, Oriol and Dahl, George E.},
date = {2017-06-12},
number = {arXiv:1704.01212},
- eprint = {1704.01212},
+ eprint = {arXiv:1704.01212},
eprinttype = {arxiv},
- primaryclass = {cs},
- publisher = {{arXiv}},
doi = {10.48550/arXiv.1704.01212},
url = {http://arxiv.org/abs/1704.01212},
urldate = {2022-10-03},
abstract = {Supervised learning on molecules has incredible potential to be useful in chemistry, drug discovery, and materials science. Luckily, several promising and closely related neural network models invariant to molecular symmetries have already been described in the literature. These models learn a message passing algorithm and aggregation procedure to compute a function of their entire input graph. At this point, the next step is to find a particularly effective variant of this general approach and apply it to chemical prediction benchmarks until we either solve them or reach the limits of the approach. In this paper, we reformulate existing models into a single common framework we call Message Passing Neural Networks (MPNNs) and explore additional novel variations within this framework. Using MPNNs we demonstrate state of the art results on an important molecular property prediction benchmark; these results are strong enough that we believe future work should focus on datasets with larger molecules or more accurate ground truth labels.},
- archiveprefix = {arXiv},
+ pubstate = {preprint},
keywords = {DeepMind,GCN,GNN,Google,ML,molecules,MPNN},
- file = {/home/johannes/Nextcloud/Zotero/Gilmer et al_2017_Neural Message Passing for Quantum Chemistry.pdf;/home/johannes/Zotero/storage/A2EV2Y8T/1704.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Gilmer et al_2017_Neural Message Passing for Quantum Chemistry.pdf;/Users/wasmer/Zotero/storage/A2EV2Y8T/1704.html}
}
@book{girvinModernCondensedMatter2019,
@@ -2844,7 +3158,7 @@
isbn = {978-1-316-48064-9},
langid = {english},
keywords = {_tablet,condensed matter,graduate,magnetism,superconductor,textbook,topological insulator},
- file = {/home/johannes/Nextcloud/Zotero/Girvin_Yang_2019_Modern Condensed Matter Physics.pdf;/home/johannes/Zotero/storage/3FP65JQ3/F0A27AC5DEA8A40EA6EA5D727ED8B14E.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Girvin_Yang_2019_Modern Condensed Matter Physics.pdf;/Users/wasmer/Zotero/storage/3FP65JQ3/F0A27AC5DEA8A40EA6EA5D727ED8B14E.html}
}
@article{glielmoAccurateInteratomicForce2017,
@@ -2862,24 +3176,22 @@
urldate = {2021-10-19},
abstract = {We present a novel scheme to accurately predict atomic forces as vector quantities, rather than sets of scalar components, by Gaussian process (GP) regression. This is based on matrix-valued kernel functions, on which we impose the requirements that the predicted force rotates with the target configuration and is independent of any rotations applied to the configuration database entries. We show that such covariant GP kernels can be obtained by integration over the elements of the rotation group SO(d) for the relevant dimensionality d. Remarkably, in specific cases the integration can be carried out analytically and yields a conservative force field that can be recast into a pair interaction form. Finally, we show that restricting the integration to a summation over the elements of a finite point group relevant to the target system is sufficient to recover an accurate GP. The accuracy of our kernels in predicting quantum-mechanical forces in real materials is investigated by tests on pure and defective Ni, Fe, and Si crystalline systems.},
keywords = {ML,nonscalar learning target,vectorial learning target},
- file = {/home/johannes/Nextcloud/Zotero/Glielmo et al_2017_Accurate interatomic force fields via machine learning with covariant kernels.pdf;/home/johannes/Zotero/storage/LVBU2R8M/Glielmo et al. - 2017 - Accurate interatomic force fields via machine lear.pdf;/home/johannes/Zotero/storage/RVCFAL4C/PhysRevB.95.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Glielmo et al_2017_Accurate interatomic force fields via machine learning with covariant kernels.pdf;/Users/wasmer/Zotero/storage/LVBU2R8M/Glielmo et al. - 2017 - Accurate interatomic force fields via machine lear.pdf;/Users/wasmer/Zotero/storage/RVCFAL4C/PhysRevB.95.html}
}
@unpublished{glielmoDADApyDistancebasedAnalysis2022,
title = {{{DADApy}}: {{Distance-based Analysis}} of {{DAta-manifolds}} in {{Python}}},
shorttitle = {{{DADApy}}},
- author = {Glielmo, Aldo and Macocco, Iuri and Doimo, Diego and Carli, Matteo and Zeni, Claudio and Wild, Romina and d' Errico, Maria and Rodriguez, Alex and Laio, Alessandro},
- options = {useprefix=true},
+ author = {Glielmo, Aldo and Macocco, Iuri and Doimo, Diego and Carli, Matteo and Zeni, Claudio and Wild, Romina and family=Errico, given=Maria, prefix=d', useprefix=true and Rodriguez, Alex and Laio, Alessandro},
date = {2022-05-04},
eprint = {2205.03373},
eprinttype = {arxiv},
- primaryclass = {physics, stat},
+ eprintclass = {physics, stat},
url = {http://arxiv.org/abs/2205.03373},
urldate = {2022-05-11},
abstract = {DADApy is a python software package for analysing and characterising high-dimensional data manifolds. It provides methods for estimating the intrinsic dimension and the probability density, for performing density-based clustering and for comparing different distance metrics. We review the main functionalities of the package and exemplify its usage in toy cases and in a real-world application. The package is freely available under the open-source Apache 2.0 license and can be downloaded from the Github page https://github.com/sissa-data-science/DADApy.},
- archiveprefix = {arXiv},
keywords = {clustering,DADApy,data exploration,density estimation,feature importance,Information imbalance,intrinsic dimension,kNN,library,Manifolds,MD,SISSA,unsupervised learning,with-code},
- file = {/home/johannes/Nextcloud/Zotero/Glielmo et al_2022_DADApy.pdf;/home/johannes/Zotero/storage/K2ZVKUHA/2205.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Glielmo et al_2022_DADApy.pdf;/Users/wasmer/Zotero/storage/K2ZVKUHA/2205.html}
}
@unpublished{glielmoRankingInformationContent2021,
@@ -2888,13 +3200,12 @@
date = {2021-04-30},
eprint = {2104.15079},
eprinttype = {arxiv},
- primaryclass = {cs, math, stat},
+ eprintclass = {cs, math, stat},
url = {http://arxiv.org/abs/2104.15079},
urldate = {2021-05-08},
abstract = {Real-world data typically contain a large number of features that are often heterogeneous in nature, relevance, and also units of measure. When assessing the similarity between data points, one can build various distance measures using subsets of these features. Using the fewest features but still retaining sufficient information about the system is crucial in many statistical learning approaches, particularly when data are sparse. We introduce a statistical test that can assess the relative information retained when using two different distance measures, and determine if they are equivalent, independent, or if one is more informative than the other. This in turn allows finding the most informative distance measure out of a pool of candidates. The approach is applied to find the most relevant policy variables for controlling the Covid-19 epidemic and to find compact yet informative representations of atomic structures, but its potential applications are wide ranging in many branches of science.},
- archiveprefix = {arXiv},
keywords = {ACSF,descriptor dimred,descriptors,descriptors analysis,ML,SOAP},
- file = {/home/johannes/Nextcloud/Zotero/Glielmo et al_2021_Ranking the information content of distance measures.pdf;/home/johannes/Zotero/storage/IHJU7M5J/2104.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Glielmo et al_2021_Ranking the information content of distance measures.pdf;/Users/wasmer/Zotero/storage/IHJU7M5J/2104.html}
}
@article{glielmoRankingInformationContent2022,
@@ -2910,7 +3221,7 @@
url = {https://www.pnas.org/doi/full/10.1093/pnasnexus/pgac039},
urldate = {2022-07-02},
keywords = {ACSF,descriptor comparison,descriptors,dimensionality reduction,GPR,information imbalance,MD,ML,SOAP},
- file = {/home/johannes/Nextcloud/Zotero/Glielmo et al_2022_Ranking the information content of distance measures.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Glielmo et al_2022_Ranking the information content of distance measures.pdf}
}
@online{GlobalMLOpsML,
@@ -2918,7 +3229,7 @@
url = {https://about.mlreef.com/blog/global-mlops-and-ml-tools-landscape},
urldate = {2021-05-13},
keywords = {ML,MLOps},
- file = {/home/johannes/Zotero/storage/LGPLQHSH/global-mlops-and-ml-tools-landscape.html}
+ file = {/Users/wasmer/Zotero/storage/LGPLQHSH/global-mlops-and-ml-tools-landscape.html}
}
@article{golzeGWCompendiumPractical2019,
@@ -2937,7 +3248,7 @@
abstract = {The GW approximation in electronic structure theory has become a widespread tool for predicting electronic excitations in chemical compounds and materials. In the realm of theoretical spectroscopy, the GW method provides access to charged excitations as measured in direct or inverse photoemission spectroscopy. The number of GW calculations in the past two decades has exploded with increased computing power and modern codes. The success of GW can be attributed to many factors: favorable scaling with respect to system size, a formal interpretation for charged excitation energies, the importance of dynamical screening in real systems, and its practical combination with other theories. In this review, we provide an overview of these formal and practical considerations. We expand, in detail, on the choices presented to the scientist performing GW calculations for the first time. We also give an introduction to the many-body theory behind GW, a review of modern applications like molecules and surfaces, and a perspective on methods which go beyond conventional GW calculations. This review addresses chemists, physicists and material scientists with an interest in theoretical spectroscopy. It is intended for newcomers to GW calculations but can also serve as an alternative perspective for experts and an up-to-date source of computational techniques.},
langid = {english},
keywords = {electronic structure theory,GW approximation,Many-body theory,PES,photoemission,physics,review},
- file = {/home/johannes/Nextcloud/Zotero/Golze et al_2019_The GW Compendium.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Golze et al_2019_The GW Compendium.pdf}
}
@unpublished{goodallRapidDiscoveryStable2022,
@@ -2946,31 +3257,28 @@
date = {2022-03-15},
eprint = {2106.11132},
eprinttype = {arxiv},
- primaryclass = {cond-mat, physics:physics},
+ eprintclass = {cond-mat, physics:physics},
url = {http://arxiv.org/abs/2106.11132},
urldate = {2022-05-09},
abstract = {A fundamental challenge in materials science pertains to elucidating the relationship between stoichiometry, stability, structure, and property. Recent advances have shown that machine learning can be used to learn such relationships, allowing the stability and functional properties of materials to be accurately predicted. However, most of these approaches use atomic coordinates as input and are thus bottle-necked by crystal structure identification when investigating novel materials. Our approach solves this bottleneck by coarse-graining the infinite search space of atomic coordinates into a combinatorially enumerable search space. The key idea is to use Wyckoff representations -- coordinate-free sets of symmetry-related positions in a crystal -- as the input to a machine learning model. Our model demonstrates exceptionally high precision in discovering new theoretically stable materials, identifying 1,569 materials that lie below the known convex hull of previously calculated materials from just 5,675 ab-initio calculations. Our approach opens up fundamental advances in computational materials discovery.},
- archiveprefix = {arXiv},
keywords = {Wyckoff representation},
- file = {/home/johannes/Nextcloud/Zotero/Goodall et al_2022_Rapid Discovery of Stable Materials by Coordinate-free Coarse Graining.pdf;/home/johannes/Zotero/storage/G7U8SY86/2106.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Goodall et al_2022_Rapid Discovery of Stable Materials by Coordinate-free Coarse Graining.pdf;/Users/wasmer/Zotero/storage/G7U8SY86/2106.html}
}
-@misc{goodallRapidDiscoveryStable2022a,
+@online{goodallRapidDiscoveryStable2022a,
title = {Rapid {{Discovery}} of {{Stable Materials}} by {{Coordinate-free Coarse Graining}}},
author = {Goodall, Rhys E. A. and Parackal, Abhijith S. and Faber, Felix A. and Armiento, Rickard and Lee, Alpha A.},
date = {2022-03-15},
number = {arXiv:2106.11132},
- eprint = {2106.11132},
+ eprint = {arXiv:2106.11132},
eprinttype = {arxiv},
- primaryclass = {cond-mat, physics:physics},
- publisher = {{arXiv}},
doi = {10.48550/arXiv.2106.11132},
url = {http://arxiv.org/abs/2106.11132},
urldate = {2022-10-03},
abstract = {A fundamental challenge in materials science pertains to elucidating the relationship between stoichiometry, stability, structure, and property. Recent advances have shown that machine learning can be used to learn such relationships, allowing the stability and functional properties of materials to be accurately predicted. However, most of these approaches use atomic coordinates as input and are thus bottle-necked by crystal structure identification when investigating novel materials. Our approach solves this bottleneck by coarse-graining the infinite search space of atomic coordinates into a combinatorially enumerable search space. The key idea is to use Wyckoff representations -- coordinate-free sets of symmetry-related positions in a crystal -- as the input to a machine learning model. Our model demonstrates exceptionally high precision in discovering new theoretically stable materials, identifying 1,569 materials that lie below the known convex hull of previously calculated materials from just 5,675 ab-initio calculations. Our approach opens up fundamental advances in computational materials discovery.},
- archiveprefix = {arXiv},
+ pubstate = {preprint},
keywords = {CGCNN,Condensed Matter - Materials Science,crystal structure,crystal symmetry,GNN,MPNN,original publication,Physics - Computational Physics,regression,Wren,Wyckoff positions,Wyckoff representation},
- file = {/home/johannes/Nextcloud/Zotero/Goodall et al_2022_Rapid Discovery of Stable Materials by Coordinate-free Coarse Graining2.pdf;/home/johannes/Zotero/storage/8I7WCWRJ/2106.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Goodall et al_2022_Rapid Discovery of Stable Materials by Coordinate-free Coarse Graining2.pdf;/Users/wasmer/Zotero/storage/8I7WCWRJ/2106.html}
}
@inproceedings{goodallWyckoffSetRegression2020,
@@ -2983,7 +3291,7 @@
eventtitle = {{{NeurIPS}}},
langid = {english},
keywords = {CGCNN,crystal structure,crystal symmetry,GNN,MPNN,original publication,regression,Wren,Wyckoff positions,Wyckoff representation},
- file = {/home/johannes/Zotero/storage/A5G9FYMS/Goodall et al_Wyckoff Set Regression for Materials Discovery.pdf}
+ file = {/Users/wasmer/Zotero/storage/A5G9FYMS/Goodall et al_Wyckoff Set Regression for Materials Discovery.pdf}
}
@online{GooglePythonStyle,
@@ -2994,7 +3302,7 @@
langid = {american},
organization = {{styleguide}},
keywords = {coding style guide,Python,software engineering},
- file = {/home/johannes/Zotero/storage/HRL7NEIR/pyguide.html}
+ file = {/Users/wasmer/Zotero/storage/HRL7NEIR/pyguide.html}
}
@book{gorelickHighPerformancePython2020,
@@ -3017,13 +3325,12 @@
date = {2021-05-18},
eprint = {2105.08717},
eprinttype = {arxiv},
- primaryclass = {physics, stat},
+ eprintclass = {physics, stat},
url = {http://arxiv.org/abs/2105.08717},
urldate = {2021-05-30},
abstract = {The input of almost every machine learning algorithm targeting the properties of matter at the atomic scale involves a transformation of the list of Cartesian atomic coordinates into a more symmetric representation. Many of these most popular representations can be seen as an expansion of the symmetrized correlations of the atom density, and differ mainly by the choice of basis. Here we discuss how to build an adaptive, optimal numerical basis that is chosen to represent most efficiently the structural diversity of the dataset at hand. For each training dataset, this optimal basis is unique, and can be computed at no additional cost with respect to the primitive basis by approximating it with splines. We demonstrate that this construction yields representations that are accurate and computationally efficient, presenting examples that involve both molecular and condensed-phase machine-learning models.},
- archiveprefix = {arXiv},
keywords = {descriptors,descriptors analysis,ML,SOAP},
- file = {/home/johannes/Nextcloud/Zotero/false;/home/johannes/Nextcloud/Zotero/Goscinski et al_2021_Optimal radial basis for density-based atomic representations.pdf;/home/johannes/Zotero/storage/HHPU5HMP/2105.html}
+ file = {/home/johannes/Nextcloud/Zotero/false;/Users/wasmer/Nextcloud/Zotero/Goscinski et al_2021_Optimal radial basis for density-based atomic representations.pdf;/Users/wasmer/Zotero/storage/HHPU5HMP/2105.html}
}
@article{goscinskiRoleFeatureSpace2021,
@@ -3043,25 +3350,39 @@
abstract = {Efficient, physically-inspired descriptors of the structure and composition of molecules and materials play a key role in the application of machine-learning techniques to atomistic simulations. The proliferation of approaches, as well as the fact that each choice of features can lead to very different behavior depending on how they are used, e.g. by introducing non-linear kernels and non-Euclidean metrics to manipulate them, makes it difficult to objectively compare different methods, and to address fundamental questions on how one feature space is related to another. In this work we introduce a framework to compare different sets of descriptors, and different ways of transforming them by means of metrics and kernels, in terms of the structure of the feature space that they induce. We define diagnostic tools to determine whether alternative feature spaces contain equivalent amounts of information, and whether the common information is substantially distorted when going from one feature space to another. We compare, in particular, representations that are built in terms of n-body correlations of the atom density, quantitatively assessing the information loss associated with the use of low-order features. We also investigate the impact of different choices of basis functions and hyperparameters of the widely used SOAP and Behler–Parrinello features, and investigate how the use of non-linear kernels, and of a Wasserstein-type metric, change the structure of the feature space in comparison to a simpler linear feature space.},
langid = {english},
keywords = {BPSF,descriptor comparison,descriptors analysis,ML,SOAP},
- file = {/home/johannes/Nextcloud/Zotero/Goscinski et al_2021_The role of feature space in atomistic learning.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Goscinski et al_2021_The role of feature space in atomistic learning.pdf}
}
-@misc{grisafiElectronicstructurePropertiesAtomcentered2022,
+@online{grisafiElectronicstructurePropertiesAtomcentered2022,
title = {Electronic-Structure Properties from Atom-Centered Predictions of the Electron Density},
author = {Grisafi, Andrea and Lewis, Alan M. and Rossi, Mariana and Ceriotti, Michele},
date = {2022-06-28},
number = {arXiv:2206.14087},
- eprint = {2206.14087},
+ eprint = {arXiv:2206.14087},
eprinttype = {arxiv},
- primaryclass = {cond-mat, physics:physics, stat},
- publisher = {{arXiv}},
doi = {10.48550/arXiv.2206.14087},
url = {http://arxiv.org/abs/2206.14087},
urldate = {2022-07-02},
abstract = {The electron density of a molecule or material has recently received major attention as a target quantity of machine-learning models. A natural choice to construct a model that yields transferable and linear-scaling predictions is to represent the scalar field using a multi-centered atomic basis analogous to that routinely used in density fitting approximations. However, the non-orthogonality of the basis poses challenges for the learning exercise, as it requires accounting for all the atomic density components at once. We devise a gradient-based approach to directly minimize the loss function of the regression problem in an optimized and highly sparse feature space. In so doing, we overcome the limitations associated with adopting an atom-centered model to learn the electron density over arbitrarily complex datasets, obtaining extremely accurate predictions. The enhanced framework is tested on 32-molecule periodic cells of liquid water, presenting enough complexity to require an optimal balance between accuracy and computational efficiency. We show that starting from the predicted density a single Kohn-Sham diagonalization step can be performed to access total energy components that carry an error of just 0.1 meV/atom with respect to the reference density functional calculations. Finally, we test our method on the highly heterogeneous QM9 benchmark dataset, showing that a small fraction of the training data is enough to derive ground-state total energies within chemical accuracy.},
- archiveprefix = {arXiv},
- keywords = {_tablet,DFT,dimensionality reduction,ML-DFT,ML-ESM,molecules,molecules & solids,prediction of electron density,QM9,SALTED,SOAP,solids},
- file = {/home/johannes/Nextcloud/Zotero/Grisafi et al_2022_Electronic-structure properties from atom-centered predictions of the electron.pdf;/home/johannes/Zotero/storage/QPHBS33I/2206.html}
+ pubstate = {preprint},
+ keywords = {_tablet,DFT,dimensionality reduction,equivariant,ML-DFT,ML-ESM,molecules,molecules & solids,prediction of electron density,QM9,SALTED,SOAP,solids},
+ file = {/Users/wasmer/Nextcloud/Zotero/Grisafi et al_2022_Electronic-structure properties from atom-centered predictions of the electron.pdf;/Users/wasmer/Zotero/storage/QPHBS33I/2206.html}
+}
+
+@article{grisafiElectronicStructurePropertiesAtomCentered2022,
+ title = {Electronic-{{Structure Properties}} from {{Atom-Centered Predictions}} of the {{Electron Density}}},
+ author = {Grisafi, Andrea and Lewis, Alan M. and Rossi, Mariana and Ceriotti, Michele},
+ date = {2022-12-01},
+ journaltitle = {Journal of Chemical Theory and Computation},
+ shortjournal = {J. Chem. Theory Comput.},
+ publisher = {{American Chemical Society}},
+ issn = {1549-9618},
+ doi = {10.1021/acs.jctc.2c00850},
+ url = {https://doi.org/10.1021/acs.jctc.2c00850},
+ urldate = {2023-01-25},
+ abstract = {The electron density of a molecule or material has recently received major attention as a target quantity of machine-learning models. A natural choice to construct a model that yields transferable and linear-scaling predictions is to represent the scalar field using a multicentered atomic basis analogous to that routinely used in density fitting approximations. However, the nonorthogonality of the basis poses challenges for the learning exercise, as it requires accounting for all the atomic density components at once. We devise a gradient-based approach to directly minimize the loss function of the regression problem in an optimized and highly sparse feature space. In so doing, we overcome the limitations associated with adopting an atom-centered model to learn the electron density over arbitrarily complex data sets, obtaining very accurate predictions using a comparatively small training set. The enhanced framework is tested on 32-molecule periodic cells of liquid water, presenting enough complexity to require an optimal balance between accuracy and computational efficiency. We show that starting from the predicted density a single Kohn–Sham diagonalization step can be performed to access total energy components that carry an error of just 0.1 meV/atom with respect to the reference density functional calculations. Finally, we test our method on the highly heterogeneous QM9 benchmark data set, showing that a small fraction of the training data is enough to derive ground-state total energies within chemical accuracy.},
+ keywords = {DFT,dimensionality reduction,equivariant,ML-DFT,ML-ESM,molecules,molecules & solids,prediction of electron density,QM9,SALTED,SOAP,solids},
+ file = {/Users/wasmer/Nextcloud/Zotero/Grisafi et al_2022_Electronic-Structure Properties from Atom-Centered Predictions of the Electron.pdf;/Users/wasmer/Zotero/storage/29HAHUDS/acs.jctc.html}
}
@article{grisafiIncorporatingLongrangePhysics2019,
@@ -3080,7 +3401,7 @@
urldate = {2022-08-16},
abstract = {The most successful and popular machine learning models of atomic-scale properties derive their transferability from a locality ansatz. The properties of a large molecule or a bulk material are written as a sum over contributions that depend on the configurations within finite atom-centered environments. The obvious downside of this approach is that it cannot capture nonlocal, nonadditive effects such as those arising due to long-range electrostatics or quantum interference. We propose a solution to this problem by introducing nonlocal representations of the system, which are remapped as feature vectors that are defined locally and are equivariant in O(3). We consider, in particular, one form that has the same asymptotic behavior as the electrostatic potential. We demonstrate that this framework can capture nonlocal, long-range physics by building a model for the electrostatic energy of randomly distributed point-charges, for the unrelaxed binding curves of charged organic molecular dimers, and for the electronic dielectric response of liquid water. By combining a representation of the system that is sensitive to long-range correlations with the transferability of an atom-centered additive model, this method outperforms current state-of-the-art machine-learning schemes and provides a conceptual framework to incorporate nonlocal physics into atomistic machine learning.},
keywords = {ML-ESM,SA-GPR},
- file = {/home/johannes/Nextcloud/Zotero/Grisafi_Ceriotti_2019_Incorporating long-range physics in atomic-scale machine learning.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Grisafi_Ceriotti_2019_Incorporating long-range physics in atomic-scale machine learning.pdf}
}
@article{grisafiMultiscaleApproachPrediction2021,
@@ -3100,7 +3421,7 @@
abstract = {Electronic nearsightedness is one of the fundamental principles that governs the behavior of condensed matter and supports its description in terms of local entities such as chemical bonds. Locality also underlies the tremendous success of machine-learning schemes that predict quantum mechanical observables – such as the cohesive energy, the electron density, or a variety of response properties – as a sum of atom-centred contributions, based on a short-range representation of atomic environments. One of the main shortcomings of these approaches is their inability to capture physical effects ranging from electrostatic interactions to quantum delocalization, which have a long-range nature. Here we show how to build a multi-scale scheme that combines in the same framework local and non-local information, overcoming such limitations. We show that the simplest version of such features can be put in formal correspondence with a multipole expansion of permanent electrostatics. The data-driven nature of the model construction, however, makes this simple form suitable to tackle also different types of delocalized and collective effects. We present several examples that range from molecular physics to surface science and biophysics, demonstrating the ability of this multi-scale approach to model interactions driven by electrostatics, polarization and dispersion, as well as the cooperative behavior of dielectric response functions.},
langid = {english},
keywords = {ML-ESM,SA-GPR},
- file = {/home/johannes/Nextcloud/Zotero/Grisafi et al_2021_Multi-scale approach for the prediction of atomic scale properties.pdf;/home/johannes/Zotero/storage/A6YSFTUU/Grisafi et al. - 2021 - Multi-scale approach for the prediction of atomic .pdf;/home/johannes/Zotero/storage/SG2EBGXV/d0sc04934d.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Grisafi et al_2021_Multi-scale approach for the prediction of atomic scale properties.pdf;/Users/wasmer/Zotero/storage/A6YSFTUU/Grisafi et al. - 2021 - Multi-scale approach for the prediction of atomic .pdf;/Users/wasmer/Zotero/storage/SG2EBGXV/d0sc04934d.html}
}
@article{grisafiSymmetryAdaptedMachineLearning2018,
@@ -3118,7 +3439,7 @@
urldate = {2021-10-19},
abstract = {Statistical learning methods show great promise in providing an accurate prediction of materials and molecular properties, while minimizing the need for computationally demanding electronic structure calculations. The accuracy and transferability of these models are increased significantly by encoding into the learning procedure the fundamental symmetries of rotational and permutational invariance of scalar properties. However, the prediction of tensorial properties requires that the model respects the appropriate geometric transformations, rather than invariance, when the reference frame is rotated. We introduce a formalism that extends existing schemes and makes it possible to perform machine learning of tensorial properties of arbitrary rank, and for general molecular geometries. To demonstrate it, we derive a tensor kernel adapted to rotational symmetry, which is the natural generalization of the smooth overlap of atomic positions kernel commonly used for the prediction of scalar properties at the atomic scale. The performance and generality of the approach is demonstrated by learning the instantaneous response to an external electric field of water oligomers of increasing complexity, from the isolated molecule to the condensed phase.},
keywords = {GPR,lambda-SOAP,library,ML,models,nonscalar learning target,original publication,SA-GPR,with-code},
- file = {/home/johannes/Nextcloud/Zotero/Grisafi et al_2018_Symmetry-Adapted Machine Learning for Tensorial Properties of Atomistic Systems.pdf;/home/johannes/Zotero/storage/JGVW5QVD/Grisafi et al. - 2018 - Symmetry-Adapted Machine Learning for Tensorial Pr.pdf;/home/johannes/Zotero/storage/KLCEZH25/PhysRevLett.120.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Grisafi et al_2018_Symmetry-Adapted Machine Learning for Tensorial Properties of Atomistic Systems.pdf;/Users/wasmer/Zotero/storage/JGVW5QVD/Grisafi et al. - 2018 - Symmetry-Adapted Machine Learning for Tensorial Pr.pdf;/Users/wasmer/Zotero/storage/KLCEZH25/PhysRevLett.120.html}
}
@article{grisafiTransferableMachineLearningModel2019,
@@ -3137,7 +3458,7 @@
urldate = {2021-10-19},
abstract = {The electronic charge density plays a central role in determining the behavior of matter at the atomic scale, but its computational evaluation requires demanding electronic-structure calculations. We introduce an atom-centered, symmetry-adapted framework to machine-learn the valence charge density based on a small number of reference calculations. The model is highly transferable, meaning it can be trained on electronic-structure data of small molecules and used to predict the charge density of larger compounds with low, linear-scaling cost. Applications are shown for various hydrocarbon molecules of increasing complexity and flexibility, and demonstrate the accuracy of the model when predicting the density on octane and octatetraene after training exclusively on butane and butadiene. This transferable, data-driven model can be used to interpret experiments, accelerate electronic structure calculations, and compute electrostatic interactions in molecules and condensed-phase systems.},
keywords = {_tablet,DFT,GPR,lambda-SOAP,library,ML,ML-DFT,ML-ESM,models,molecules,prediction of electron density,SA-GPR,SALTED,SOAP,with-code},
- file = {/home/johannes/Nextcloud/Zotero/Grisafi et al_2019_Transferable Machine-Learning Model of the Electron Density.pdf;/home/johannes/Zotero/storage/HMBCGARZ/acscentsci.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Grisafi et al_2019_Transferable Machine-Learning Model of the Electron Density.pdf;/Users/wasmer/Zotero/storage/HMBCGARZ/acscentsci.html}
}
@book{grusDataScienceScratch2019,
@@ -3166,25 +3487,23 @@
abstract = {Background: Research results in artificial intelligence (AI) are criticized for not being reproducible. Objective: To quantify the state of reproducibility of empirical AI research using six reproducibility metrics measuring three different degrees of reproducibility. Hypotheses: 1) AI research is not documented well enough to reproduce the reported results. 2) Documentation practices have improved over time. Method: The literature is reviewed and a set of variables that should be documented to enable reproducibility are grouped into three factors: Experiment, Data and Method. The metrics describe how well the factors have been documented for a paper. A total of 400 research papers from the conference series IJCAI and AAAI have been surveyed using the metrics. Findings: None of the papers document all of the variables. The metrics show that between 20\% and 30\% of the variables for each factor are documented. One of the metrics show statistically significant increase over time while the others show no change. Interpretation: The reproducibility scores decrease with in- creased documentation requirements. Improvement over time is found. Conclusion: Both hypotheses are supported.},
eventtitle = {Thirty-{{Second AAAI Conference}} on {{Artificial Intelligence}}},
langid = {english},
- file = {/home/johannes/Nextcloud/Zotero/Gundersen_Kjensmo_2018_State of the Art.pdf;/home/johannes/Zotero/storage/PXBAF83Q/17248.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Gundersen_Kjensmo_2018_State of the Art.pdf;/Users/wasmer/Zotero/storage/PXBAF83Q/17248.html}
}
-@misc{gutmannPenPaperExercises2022,
+@online{gutmannPenPaperExercises2022,
title = {Pen and {{Paper Exercises}} in {{Machine Learning}}},
author = {Gutmann, Michael U.},
date = {2022-06-27},
number = {arXiv:2206.13446},
- eprint = {2206.13446},
+ eprint = {arXiv:2206.13446},
eprinttype = {arxiv},
- primaryclass = {cs, stat},
- publisher = {{arXiv}},
doi = {10.48550/arXiv.2206.13446},
url = {http://arxiv.org/abs/2206.13446},
urldate = {2022-06-29},
abstract = {This is a collection of (mostly) pen-and-paper exercises in machine learning. The exercises are on the following topics: linear algebra, optimisation, directed graphical models, undirected graphical models, expressive power of graphical models, factor graphs and message passing, inference for hidden Markov models, model-based learning (including ICA and unnormalised models), sampling and Monte-Carlo integration, and variational inference.},
- archiveprefix = {arXiv},
+ pubstate = {preprint},
keywords = {course material,exercises,General ML,graphical model,hidden Markov model,MC integration,ML,sampling,teaching},
- file = {/home/johannes/Nextcloud/Zotero/Gutmann_2022_Pen and Paper Exercises in Machine Learning.pdf;/home/johannes/Zotero/storage/KMSFX6RY/2206.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Gutmann_2022_Pen and Paper Exercises in Machine Learning.pdf;/Users/wasmer/Zotero/storage/KMSFX6RY/2206.html}
}
@article{hafizHighthroughputDataAnalysis2018,
@@ -3204,7 +3523,7 @@
issue = {1},
langid = {english},
keywords = {HTC,materials informatics,strongly correlated maeterials},
- file = {/home/johannes/Nextcloud/Zotero/Hafiz et al_2018_A high-throughput data analysis and materials discovery tool for strongly.pdf;/home/johannes/Zotero/storage/VAW7UFBH/s41524-018-0120-9.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Hafiz et al_2018_A high-throughput data analysis and materials discovery tool for strongly.pdf;/Users/wasmer/Zotero/storage/VAW7UFBH/s41524-018-0120-9.html}
}
@article{handleyNextGenerationInteratomic2014,
@@ -3223,7 +3542,7 @@
abstract = {The computer simulation of condensed systems is a challenging task. While electronic structure methods like density-functional theory (DFT) usually provide a good compromise between accuracy and efficiency, they are computationally very demanding and thus applicable only to systems containing up to a few hundred atoms. Unfortunately, many interesting problems require simulations to be performed on much larger systems involving thousands of atoms or more. Consequently, more efficient methods are urgently needed, and a lot of effort has been spent on the development of a large variety of potentials enabling simulations with significantly extended time and length scales. Most commonly, these potentials are based on physically motivated functional forms and thus perform very well for the applications they have been designed for. On the other hand, they are often highly system-specific and thus cannot easily be transferred from one system to another. Moreover, their numerical accuracy is restricted by the intrinsic limitations of the imposed functional forms. In recent years, several novel types of potentials have emerged, which are not based on physical considerations. Instead, they aim to reproduce a set of reference electronic structure data as accurately as possible by using very general and flexible functional forms. In this review we will survey a number of these methods. While they differ in the choice of the employed mathematical functions, they all have in common that they provide high-quality potential-energy surfaces, while the efficiency is comparable to conventional empirical potentials. It has been demonstrated that in many cases these potentials now offer a very interesting new approach to study complex systems with hitherto unreached accuracy.},
langid = {english},
keywords = {condensed,ML,MLP,models},
- file = {/home/johannes/Nextcloud/Zotero/Handley_Behler_2014_Next generation interatomic potentials for condensed systems.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Handley_Behler_2014_Next generation interatomic potentials for condensed systems.pdf}
}
@article{hartlNationaleForschungsdateninfrastrukturNFDI2021,
@@ -3241,7 +3560,7 @@
urldate = {2021-10-15},
abstract = {In der Nationalen Forschungsdateninfrastruktur (NFDI) werden wertvolle Forschungsdaten für das gesamte deutsche Wissenschaftssystem systematisch erschlossen, vernetzt und nachhaltig nutzbar gemacht. Bislang sind diese meist dezentral, projektbezogen oder nur zeitlich begrenzt verfügbar. Mit der NFDI soll ein digitaler Wissensspeicher unter Berücksichtigung der FAIR-Prinzipien (Findable, Accessible, Interoperable, Reusable) geschaffen werden. Bereits vorhandene Daten können zur Bearbeitung weiterer Forschungsfragen genutzt werden und neue Erkenntnisse und Innovationen ermöglichen.Bis zu 30 NFDI-Konsortien, Zusammenschlüsse verschiedener Einrichtungen innerhalb eines Forschungsfeldes, arbeiten zusammen interdisziplinär an der Zielumsetzung. Um die Aktivitäten zum Aufbau einer Nationalen Forschungsdateninfrastruktur zu koordinieren, wurde der gemeinnützige Verein Nationale Forschungsdateninfrastruktur (NFDI) e.V. mit Sitz in Karlsruhe im Oktober 2020 gegründet. Gemeinsam gestalten Verein und NFDI-Konsortien die Zukunft des Forschungsdatenmanagements in Deutschland. Darüber hinaus soll NFDI auch am Aufbau internationaler Initiativen, beispielsweise der European Open Science Cloud (EOSC), mitwirken.},
langid = {ngerman},
- file = {/home/johannes/Nextcloud/Zotero/Hartl et al_2021_Nationale Forschungsdateninfrastruktur (NFDI).pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Hartl et al_2021_Nationale Forschungsdateninfrastruktur (NFDI).pdf}
}
@article{hartmaierDataOrientedConstitutiveModeling2020,
@@ -3261,7 +3580,7 @@
issue = {7},
langid = {english},
keywords = {constitutive modeling,FEM,ML},
- file = {/home/johannes/Nextcloud/Zotero/Hartmaier_2020_Data-Oriented Constitutive Modeling of Plasticity in Metals.pdf;/home/johannes/Zotero/storage/94LTSN79/htm.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Hartmaier_2020_Data-Oriented Constitutive Modeling of Plasticity in Metals.pdf;/Users/wasmer/Zotero/storage/94LTSN79/htm.html}
}
@article{hasnipDensityFunctionalTheory2014,
@@ -3278,7 +3597,7 @@
urldate = {2022-05-18},
abstract = {Density functional theory (DFT) has been used in many fields of the physical sciences, but none so successfully as in the solid state. From its origins in condensed matter physics, it has expanded into materials science, high-pressure physics and mineralogy, solid-state chemistry and more, powering entire computational subdisciplines. Modern DFT simulation codes can calculate a vast range of structural, chemical, optical, spectroscopic, elastic, vibrational and thermodynamic phenomena. The ability to predict structure–property relationships has revolutionized experimental fields, such as vibrational and solid-state NMR spectroscopy, where it is the primary method to analyse and interpret experimental spectra. In semiconductor physics, great progress has been made in the electronic structure of bulk and defect states despite the severe challenges presented by the description of excited states. Studies are no longer restricted to known crystallographic structures. DFT is increasingly used as an exploratory tool for materials discovery and computational experiments, culminating in ex nihilo crystal structure prediction, which addresses the long-standing difficult problem of how to predict crystal structure polymorphs from nothing but a specified chemical composition. We present an overview of the capabilities of solid-state DFT simulations in all of these topics, illustrated with recent examples using the CASTEP computer program.},
keywords = {_tablet,condensed matter,DFT,review},
- file = {/home/johannes/Nextcloud/Zotero/Hasnip et al_2014_Density functional theory in the solid state.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Hasnip et al_2014_Density functional theory in the solid state.pdf}
}
@article{helfrechtStructurepropertyMapsKernel2020,
@@ -3296,8 +3615,8 @@
urldate = {2021-10-19},
abstract = {Data analyses based on linear methods constitute the simplest, most robust, and transparent approaches to the automatic processing of large amounts of data for building supervised or unsupervised machine learning models. Principal covariates regression (PCovR) is an underappreciated method that interpolates between principal component analysis and linear regression and can be used conveniently to reveal structure-property relations in terms of simple-to-interpret, low-dimensional maps. Here we provide a pedagogic overview of these data analysis schemes, including the use of the kernel trick to introduce an element of non-linearity while maintaining most of the convenience and the simplicity of linear approaches. We then introduce a kernelized version of PCovR and a sparsified extension, and demonstrate the performance of this approach in revealing and predicting structure-property relations in chemistry and materials science, showing a variety of examples including elemental carbon, porous silicate frameworks, organic molecules, amino acid conformers, and molecular materials.},
langid = {english},
- keywords = {KPCovR,models,original publication,PCovR},
- file = {/home/johannes/Nextcloud/Zotero/Helfrecht et al_2020_Structure-property maps with Kernel principal covariates regression.pdf}
+ keywords = {_tablet,KPCovR,KRR,models,original publication,PCovR,regression,Supervised learning,unsupervised learning},
+ file = {/Users/wasmer/Nextcloud/Zotero/Helfrecht et al_2020_Structure-property maps with Kernel principal covariates regression.pdf}
}
@article{henkTopologicalCharacterMagnetism2012,
@@ -3314,7 +3633,7 @@
url = {https://link.aps.org/doi/10.1103/PhysRevLett.109.076801},
urldate = {2022-05-13},
abstract = {First-principles and model calculations show that the Dirac surface state of the topological insulator Bi2Te3 survives upon moderate Mn doping of the surface layers but can lose its topological character as a function of magnetization direction. The dispersion depends considerably on the direction of the Mn magnetization: for perpendicular magnetization, a gap of 16 meV opens up at the Dirac point; for in-plane magnetization, a tiny gap can be opened or closed in dependence on the magnetization azimuth. The ground state is ferromagnetic, with a critical temperature of 12 K. The results provide a path towards a magnetic control of the topological character of the Dirac surface state and its consequences to spin-dependent transport properties.},
- file = {/home/johannes/Nextcloud/Zotero/Henk et al_2012_Topological Character and Magnetism of the Dirac State in Mn-Doped.pdf;/home/johannes/Zotero/storage/W6BV33VI/Henk et al. - 2012 - Topological Character and Magnetism of the Dirac S.pdf;/home/johannes/Zotero/storage/ZTFJBVIM/PhysRevLett.109.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Henk et al_2012_Topological Character and Magnetism of the Dirac State in Mn-Doped.pdf;/Users/wasmer/Zotero/storage/W6BV33VI/Henk et al. - 2012 - Topological Character and Magnetism of the Dirac S.pdf;/Users/wasmer/Zotero/storage/ZTFJBVIM/PhysRevLett.109.html}
}
@article{herbstDFTKJulianApproach2021,
@@ -3325,7 +3644,7 @@
journaltitle = {JuliaCon Proceedings},
doi = {10.21105/JCON.00069},
abstract = {Density-functional theory (DFT) is a widespread method for simulating the quantum-chemical behaviour of electrons in matter. It provides a first-principles description of many optical, mechanical and chemical properties at an acceptable computational cost [16, 2, 3]. For a wide range of systems the obtained predictions are accurate and shortcomings of the theory are by now wellunderstood [2, 3]. The desire to tackle even bigger systems and more involved materials, however, keeps posing novel challenges that require methods to constantly improve. One example are socalled high-throughput screening approaches, which are becoming prominent in recent years. In these techniques one wishes to systematically scan over huge design spaces of compounds in order to identify promising novel materials for targeted follow-up investigation. This has already lead to many success stories [14], such as the discovery of novel earth-abundant semiconductors [11], novel light-absorbing materials [20], electrocatalysts [8], materials for hydrogen storage [13] or for Li-ion batteries [1]. Keeping in mind the large range of physics that needs to be covered in these studies as well as the typical number of calculations (up to the order of millions), a bottleneck in these studies is the reliability and performance of the underlying DFT codes.},
- file = {/home/johannes/Nextcloud/Zotero/Herbst et al_2021_DFTK.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Herbst et al_2021_DFTK.pdf}
}
@unpublished{herbstSurrogateModelsQuantum2021,
@@ -3334,13 +3653,12 @@
date = {2021-11-24},
eprint = {2110.15665},
eprinttype = {arxiv},
- primaryclass = {cond-mat, physics:quant-ph},
+ eprintclass = {cond-mat, physics:quant-ph},
url = {http://arxiv.org/abs/2110.15665},
urldate = {2022-01-11},
abstract = {We present a methodology to investigate phase-diagrams of quantum models based on the principle of the reduced basis method (RBM). The RBM is built from a few ground-state snapshots, i.e., lowest eigenvectors of the full system Hamiltonian computed at well-chosen points in the parameter space of interest. We put forward a greedy-strategy to assemble such small-dimensional basis, i.e., to select where to spend the numerical effort needed for the snapshots. Once the RBM is assembled, physical observables required for mapping out the phase-diagram (e.g., structure factors) can be computed for any parameter value with a modest computational complexity, considerably lower than the one associated to the underlying Hilbert space dimension. We benchmark the method in two test cases, a chain of excited Rydberg atoms and a geometrically frustrated antiferromagnetic two-dimensional lattice model, and illustrate the accuracy of the approach. In particular, we find that the ground-manifold can be approximated to sufficient accuracy with a moderate number of basis functions, which increases very mildly when the number of microscopic constituents grows - in stark contrast to the exponential growth of the Hilbert space needed to describe each of the few snapshots. A combination of the presented RBM approach with other numerical techniques circumventing even the latter big cost, e.g., Tensor Network methods, is a tantalising outlook of this work.},
- archiveprefix = {arXiv},
keywords = {Condensed Matter - Strongly Correlated Electrons,Quantum Physics},
- file = {/home/johannes/Nextcloud/Zotero/Herbst et al_2021_Surrogate models for quantum spin systems based on reduced order modeling.pdf;/home/johannes/Zotero/storage/F4FW6AHT/2110.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Herbst et al_2021_Surrogate models for quantum spin systems based on reduced order modeling.pdf;/Users/wasmer/Zotero/storage/F4FW6AHT/2110.html}
}
@article{hermannDeepneuralnetworkSolutionElectronic2020,
@@ -3361,7 +3679,7 @@
issue = {10},
langid = {english},
keywords = {Computational chemistry,Method development,Physical chemistry,Quantum chemistry,Theoretical chemistry},
- file = {/home/johannes/Nextcloud/Zotero/Hermann et al_2020_Deep-neural-network solution of the electronic Schrödinger equation.pdf;/home/johannes/Zotero/storage/V947YTSM/s41557-020-0544-y.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Hermann et al_2020_Deep-neural-network solution of the electronic Schrödinger equation.pdf;/Users/wasmer/Zotero/storage/V947YTSM/s41557-020-0544-y.html}
}
@article{herrCompressingPhysicsAutoencoder2019,
@@ -3379,7 +3697,7 @@
doi = {10.1063/1.5108803},
url = {https://aip.scitation.org/doi/full/10.1063/1.5108803},
urldate = {2022-01-02},
- file = {/home/johannes/Nextcloud/Zotero/Herr et al_2019_Compressing physics with an autoencoder.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Herr et al_2019_Compressing physics with an autoencoder.pdf}
}
@article{himanenDScribeLibraryDescriptors2020,
@@ -3398,7 +3716,7 @@
abstract = {DScribe is a software package for machine learning that provides popular feature transformations (“descriptorsâ€) for atomistic materials simulations. DScribe accelerates the application of machine learning for atomistic property prediction by providing user-friendly, off-the-shelf descriptor implementations. The package currently contains implementations for Coulomb matrix, Ewald sum matrix, sine matrix, Many-body Tensor Representation (MBTR), Atom-centered Symmetry Function (ACSF) and Smooth Overlap of Atomic Positions (SOAP). Usage of the package is illustrated for two different applications: formation energy prediction for solids and ionic charge prediction for atoms in organic molecules. The package is freely available under the open-source Apache License 2.0. Program summary Program Title: DScribe Program Files doi: http://dx.doi.org/10.17632/vzrs8n8pk6.1 Licensing provisions: Apache-2.0 Programming language: Python/C/C++ Supplementary material: Supplementary Information as PDF Nature of problem: The application of machine learning for materials science is hindered by the lack of consistent software implementations for feature transformations. These feature transformations, also called descriptors, are a key step in building machine learning models for property prediction in materials science. Solution method: We have developed a library for creating common descriptors used in machine learning applied to materials science. We provide an implementation the following descriptors: Coulomb matrix, Ewald sum matrix, sine matrix, Many-body Tensor Representation (MBTR), Atom-centered Symmetry Functions (ACSF) and Smooth Overlap of Atomic Positions (SOAP). The library has a python interface with computationally intensive routines written in C or C++. The source code, tutorials and documentation are provided online. A continuous integration mechanism is set up to automatically run a series of regression tests and check code coverage when the codebase is updated.},
langid = {english},
keywords = {ACSF,descriptors,DScribe,library,materials,Matrix descriptors,MBTR,ML,Open source,Python,rec-by-ruess,SOAP,with-code},
- file = {/home/johannes/Nextcloud/Zotero/Himanen et al_2020_DScribe_Supplementary.pdf;/home/johannes/Nextcloud/Zotero/Himanen et al_2020_DScribe.pdf;/home/johannes/Zotero/storage/IZ66FYIN/S0010465519303042.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Himanen et al_2020_DScribe_Supplementary.pdf;/Users/wasmer/Nextcloud/Zotero/Himanen et al_2020_DScribe.pdf;/Users/wasmer/Zotero/storage/IZ66FYIN/S0010465519303042.html}
}
@book{himanenMaterialsInformaticsAugmenting2020,
@@ -3414,7 +3732,7 @@
langid = {english},
keywords = {DScribe,ML,review},
annotation = {Accepted: 2020-02-11T10:01:06Z},
- file = {/home/johannes/Nextcloud/Zotero/Himanen_2020_Materials Informatics - Augmenting Materials Research with Data-driven Design.pdf;/home/johannes/Zotero/storage/5N3DHF4G/43027.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Himanen_2020_Materials Informatics - Augmenting Materials Research with Data-driven Design.pdf;/Users/wasmer/Zotero/storage/5N3DHF4G/43027.html}
}
@article{hirohataReviewSpintronicsPrinciples2020,
@@ -3433,7 +3751,7 @@
abstract = {Spintronics is one of the emerging fields for the next-generation nanoelectronic devices to reduce their power consumption and to increase their memory and processing capabilities. Such devices utilise the spin degree of freedom of electrons and/or holes, which can also interact with their orbital moments. In these devices, the spin polarisation is controlled either by magnetic layers used as spin-polarisers or analysers or via spin–orbit coupling. Spin waves can also be used to carry spin current. In this review, the fundamental physics of these phenomena is described first with respect to the spin generation methods as detailed in Sections 2~\textasciitilde ~9. The recent development in their device applications then follows in Sections 10 and 11. Future perspectives are provided at the end.},
langid = {english},
keywords = {Dzyaloshinskii–Moriya interaction,Electric field,Electromagnetic wave,Hard disk drive,Landau-Lifshits-Gilbert equation,Magnetic damping,Magnetic random access memory,Magnetic sensor,Magnetic skyrmion,Neuromorphic,Racetrack memory,Spin Hall effects,Spin Nernst effect,Spin Seebeck effect,Spin-current generation,Spin-orbit effects,Spin-transfer torque,Spintronics},
- file = {/home/johannes/Nextcloud/Zotero/Hirohata et al_2020_Review on spintronics.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Hirohata et al_2020_Review on spintronics.pdf}
}
@article{hohenbergInhomogeneousElectronGas1964,
@@ -3447,7 +3765,7 @@
doi = {10.1103/PhysRev.136.B864},
issue = {3B},
keywords = {DFT,HKT,original publication},
- file = {/home/johannes/Nextcloud/Zotero/Hohenberg_1964_Inhomogeneous Electron Gas.pdf;/home/johannes/Zotero/storage/BRS4FL49/PhysRev.136.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Hohenberg_1964_Inhomogeneous Electron Gas.pdf;/Users/wasmer/Zotero/storage/BRS4FL49/PhysRev.136.html}
}
@article{holecAtomisticModelingBasedDesign2017,
@@ -3465,8 +3783,7 @@
abstract = {Modern materials science increasingly advances via a knowledge-based development rather than a trial-and-error procedure. Gathering large amounts of data and getting deep understanding of non-trivial relationships between synthesis of materials, their structure and properties is experimentally a tedious work. Here, theoretical modeling plays a vital role. In this review paper we briefly introduce modeling approaches employed in materials science, their principles and fields of application. We then focus on atomistic modeling methods, mostly quantum mechanical ones but also Monte Carlo and classical molecular dynamics, to demonstrate their practical use on selected examples.},
langid = {english},
keywords = {materials,mechanical,MPI Eisenforschung,multiscale},
- annotation = {\_eprint: https://onlinelibrary.wiley.com/doi/pdf/10.1002/adem.201600688},
- file = {/home/johannes/Nextcloud/Zotero/Holec et al_2017_Atomistic Modeling-Based Design of Novel Materials.pdf;/home/johannes/Zotero/storage/NHGDI7UK/adem.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Holec et al_2017_Atomistic Modeling-Based Design of Novel Materials.pdf;/Users/wasmer/Zotero/storage/NHGDI7UK/adem.html}
}
@article{hollingsworthCanExactConditions2018,
@@ -3485,7 +3802,7 @@
urldate = {2022-07-08},
abstract = {Historical methods of functional development in density functional theory have often been guided by analytic conditions that constrain the exact functional one is trying to approximate. Recently, machine-learned functionals have been created by interpolating the results from a small number of exactly solved systems to unsolved systems that are similar in nature. For a simple one-dimensional system, using an exact condition, we find improvements in the learning curves of a machine learning approximation to the non-interacting kinetic energy functional. We also find that the significance of the improvement depends on the nature of the interpolation manifold of the machine-learned functional.},
keywords = {DFA,DFT,KRR,ML,ML-DFA,ML-DFT,ML-ESM},
- file = {/home/johannes/Nextcloud/Zotero/Hollingsworth et al_2018_Can exact conditions improve machine-learned density functionals.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Hollingsworth et al_2018_Can exact conditions improve machine-learned density functionals.pdf}
}
@book{hollPhysicsbasedDeepLearning2021,
@@ -3497,7 +3814,7 @@
urldate = {2022-10-02},
abstract = {Welcome to the Physics-based Deep Learning Book (v0.2) 👋 TL;DR: This document contains a practical and comprehensive introduction of everything related to deep learning in the context of physical simulations. As much as possible, all topics come with hands-on code examples in the form of Jupyter notebooks to quickly get started. Beyond standard supervised learning from data, we’ll look at physical loss constraints, more tightly coupled learning algorithms with differentiable simulations, training algorithms tailored to physics problems, as well as reinforcement learning and uncertainty modeling. We live in exciting times: these methods have a huge potential to fundamentally change what computer simulations can achieve.},
keywords = {autodiff,basics,book,Deep learning,learning material,ML,online book,physics-based deep learning,physics-informed ML,PINN},
- file = {/home/johannes/Zotero/storage/Q7HV8D2L/intro.html}
+ file = {/Users/wasmer/Zotero/storage/Q7HV8D2L/intro.html}
}
@article{hongReducingTimeDiscovery2021,
@@ -3517,7 +3834,7 @@
abstract = {Multiscale and multimodal imaging of material structures and properties provides solid ground on which materials theory and design can flourish. Recently, KAIST announced 10 flagship research fields, which include KAIST Materials Revolution: Materials and Molecular Modeling, Imaging, Informatics and Integration (M3I3). The M3I3 initiative aims to reduce the time for the discovery, design and development of materials based on elucidating multiscale processing-structure-property relationship and materials hierarchy, which are to be quantified and understood through a combination of machine learning and scientific insights. In this review, we begin by introducing recent progress on related initiatives around the globe, such as the Materials Genome Initiative (U.S.), Materials Informatics (U.S.), the Materials Project (U.S.), the Open Quantum Materials Database (U.S.), Materials Research by Information Integration Initiative (Japan), Novel Materials Discovery (E.U.), the NOMAD repository (E.U.), Materials Scientific Data Sharing Network (China), Vom Materials Zur Innovation (Germany), and Creative Materials Discovery (Korea), and discuss the role of multiscale materials and molecular imaging combined with machine learning in realizing the vision of M3I3. Specifically, microscopies using photons, electrons, and physical probes will be revisited with a focus on the multiscale structural hierarchy, as well as structure-property relationships. Additionally, data mining from the literature combined with machine learning will be shown to be more efficient in finding the future direction of materials structures with improved properties than the classical approach. Examples of materials for applications in energy and information will be reviewed and discussed. A case study on the development of a Ni-Co-Mn cathode materials illustrates M3I3's approach to creating libraries of multiscale structure-property-processing relationships. We end with a future outlook toward recent developments in the field of M3I3.},
langid = {english},
keywords = {KAIST,Li-ion battery,M3I3,machine learning,materials and molecular modeling,materials imaging,materials informatics,materials integration},
- file = {/home/johannes/Nextcloud/Zotero/Hong et al_2021_Reducing Time to Discovery.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Hong et al_2021_Reducing Time to Discovery.pdf}
}
@article{huberAiiDAScalableComputational2020,
@@ -3539,7 +3856,7 @@
langid = {english},
keywords = {AiiDA,original publication},
annotation = {Bandiera\_abtest: a Cc\_license\_type: cc\_by Cg\_type: Nature Research Journals Primary\_atype: Research Subject\_term: Computational methods;Research management Subject\_term\_id: computational-methods;research-management},
- file = {/home/johannes/Nextcloud/Zotero/Huber et al_2020_AiiDA 1.pdf;/home/johannes/Zotero/storage/SQ25VE8T/s41597-020-00638-4.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Huber et al_2020_AiiDA 1.pdf;/Users/wasmer/Zotero/storage/SQ25VE8T/s41597-020-00638-4.html}
}
@unpublished{huberCommonWorkflowsComputing2021,
@@ -3548,13 +3865,33 @@
date = {2021-05-11},
eprint = {2105.05063},
eprinttype = {arxiv},
- primaryclass = {cond-mat},
+ eprintclass = {cond-mat},
url = {http://arxiv.org/abs/2105.05063},
urldate = {2021-06-23},
abstract = {The prediction of material properties through electronic-structure simulations based on density-functional theory has become routinely common, thanks, in part, to the steady increase in the number and robustness of available simulation packages. This plurality of codes and methods aiming to solve similar problems is both a boon and a burden. While providing great opportunities for cross-verification, these packages adopt different methods, algorithms, and paradigms, making it challenging to choose, master, and efficiently use any one for a given task. Leveraging recent advances in managing reproducible scientific workflows, we demonstrate how developing common interfaces for workflows that automatically compute material properties can tackle the challenge mentioned above, greatly simplifying interoperability and cross-verification. We introduce design rules for reproducible and reusable code-agnostic workflow interfaces to compute well-defined material properties, which we implement for eleven different quantum engines and use to compute three different material properties. Each implementation encodes carefully selected simulation parameters and workflow logic, making the implementer's expertise of the quantum engine directly available to non-experts. Full provenance and reproducibility of the workflows is guaranteed through the use of the AiiDA infrastructure. All workflows are made available as open-source and come pre-installed with the Quantum Mobile virtual machine, making their use straightforward.},
- archiveprefix = {arXiv},
keywords = {AiiDA,workflows},
- file = {/home/johannes/Nextcloud/Zotero/Huber et al_2021_Common workflows for computing material properties using different quantum.pdf;/home/johannes/Zotero/storage/9AJRBXBR/2105.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Huber et al_2021_Common workflows for computing material properties using different quantum.pdf;/Users/wasmer/Zotero/storage/9AJRBXBR/2105.html}
+}
+
+@article{huberCommonWorkflowsComputing2021a,
+ title = {Common Workflows for Computing Material Properties Using Different Quantum Engines},
+ author = {Huber, Sebastiaan P. and Bosoni, Emanuele and Bercx, Marnik and Bröder, Jens and Degomme, Augustin and Dikan, Vladimir and Eimre, Kristjan and Flage-Larsen, Espen and Garcia, Alberto and Genovese, Luigi and Gresch, Dominik and Johnston, Conrad and Petretto, Guido and Poncé, Samuel and Rignanese, Gian-Marco and Sewell, Christopher J. and Smit, Berend and Tseplyaev, Vasily and Uhrin, Martin and Wortmann, Daniel and Yakutovich, Aliaksandr V. and Zadoks, Austin and Zarabadi-Poor, Pezhman and Zhu, Bonan and Marzari, Nicola and Pizzi, Giovanni},
+ date = {2021-08-19},
+ journaltitle = {npj Computational Materials},
+ shortjournal = {npj Comput Mater},
+ volume = {7},
+ number = {1},
+ pages = {1--12},
+ publisher = {{Nature Publishing Group}},
+ issn = {2057-3960},
+ doi = {10.1038/s41524-021-00594-6},
+ url = {https://www.nature.com/articles/s41524-021-00594-6},
+ urldate = {2023-03-13},
+ abstract = {The prediction of material properties based on density-functional theory has become routinely common, thanks, in part, to the steady increase in the number and robustness of available simulation packages. This plurality of codes and methods is both a boon and a burden. While providing great opportunities for cross-verification, these packages adopt different methods, algorithms, and paradigms, making it challenging to choose, master, and efficiently use them. We demonstrate how developing common interfaces for workflows that automatically compute material properties greatly simplifies interoperability and cross-verification. We introduce design rules for reusable, code-agnostic, workflow interfaces to compute well-defined material properties, which we implement for eleven quantum engines and use to compute various material properties. Each implementation encodes carefully selected simulation parameters and workflow logic, making the implementer’s expertise of the quantum engine directly available to non-experts. All workflows are made available as open-source and full reproducibility of the workflows is guaranteed through the use of the AiiDA infrastructure.},
+ issue = {1},
+ langid = {english},
+ keywords = {/unread,AiiDA,AiiDA-FLEUR,DFT,FAIR,FLEUR,provenance,workflows},
+ file = {/Users/wasmer/Nextcloud/Zotero/Huber et al_2021_Common workflows for computing material properties using different quantum3.pdf}
}
@unpublished{huoUnifiedRepresentationMolecules2018,
@@ -3563,13 +3900,12 @@
date = {2018-01-02},
eprint = {1704.06439},
eprinttype = {arxiv},
- primaryclass = {cond-mat, physics:physics},
+ eprintclass = {cond-mat, physics:physics},
url = {http://arxiv.org/abs/1704.06439},
urldate = {2021-06-29},
abstract = {Accurate simulations of atomistic systems from first principles are limited by computational cost. In high-throughput settings, machine learning can potentially reduce these costs significantly by accurately interpolating between reference calculations. For this, kernel learning approaches crucially require a single Hilbert space accommodating arbitrary atomistic systems. We introduce a many-body tensor representation that is invariant to translations, rotations and nuclear permutations of same elements, unique, differentiable, can represent molecules and crystals, and is fast to compute. Empirical evidence is presented for energy prediction errors below 1 kcal/mol for 7k organic molecules and 5 meV/atom for 11k elpasolite crystals. Applicability is demonstrated for phase diagrams of Pt-group/transition-metal binary systems.},
- archiveprefix = {arXiv},
keywords = {descriptors,MBTR,ML,original publication},
- file = {/home/johannes/Nextcloud/Zotero/Huo_Rupp_2018_Unified Representation of Molecules and Crystals for Machine Learning.pdf;/home/johannes/Zotero/storage/EZJ986KS/1704.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Huo_Rupp_2018_Unified Representation of Molecules and Crystals for Machine Learning.pdf;/Users/wasmer/Zotero/storage/EZJ986KS/1704.html}
}
@article{hutsonArtificialIntelligenceFaces2018,
@@ -3604,7 +3940,7 @@
abstract = {Data exploration is about efficiently extracting knowledge from data even if we do not know exactly what we are looking for. In this tutorial, we survey recent developments in the emerging area of database systems tailored for data exploration. We discuss new ideas on how to store and access data as well as new ideas on how to interact with a data system to enable users and applications to quickly figure out which data parts are of interest. In addition, we discuss how to exploit lessons-learned from past research, the new challenges data exploration crafts, emerging applications and future research directions.},
isbn = {978-1-4503-2758-9},
keywords = {data exploration,Database,unsupervised learning,visualization},
- file = {/home/johannes/Nextcloud/Zotero/Idreos et al_2015_Overview of Data Exploration Techniques.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Idreos et al_2015_Overview of Data Exploration Techniques.pdf}
}
@article{imbalzanoAutomaticSelectionAtomic2018,
@@ -3623,7 +3959,7 @@
urldate = {2021-05-18},
abstract = {Machine learning of atomic-scale properties is revolutionizing molecular modeling, making it possible to evaluate inter-atomic potentials with first-principles accuracy, at a fraction of the costs. The accuracy, speed, and reliability of machine learning potentials, however, depend strongly on the way atomic configurations are represented, i.e., the choice of descriptors used as input for the machine learning method. The raw Cartesian coordinates are typically transformed in “fingerprints,†or “symmetry functions,†that are designed to encode, in addition to the structure, important properties of the potential energy surface like its invariances with respect to rotation, translation, and permutation of like atoms. Here we discuss automatic protocols to select a number of fingerprints out of a large pool of candidates, based on the correlations that are intrinsic to the training data. This procedure can greatly simplify the construction of neural network potentials that strike the best balance between accuracy and computational efficiency and has the potential to accelerate by orders of magnitude the evaluation of Gaussian approximation potentials based on the smooth overlap of atomic positions kernel. We present applications to the construction of neural network potentials for water and for an Al–Mg–Si alloy and to the prediction of the formation energies of small organic molecules using Gaussian process regression.},
keywords = {ACSF,autoML,descriptors,GAP,ML,MLP},
- file = {/home/johannes/Nextcloud/Zotero/Imbalzano et al_2018_Automatic selection of atomic fingerprints and reference configurations for.pdf;/home/johannes/Zotero/storage/DBZXGYRI/1.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Imbalzano et al_2018_Automatic selection of atomic fingerprints and reference configurations for.pdf;/Users/wasmer/Zotero/storage/DBZXGYRI/1.html}
}
@online{ImprovingDensityFunctional,
@@ -3633,7 +3969,7 @@
abstract = {Explore millions of resources from scholarly journals, books, newspapers, videos and more, on the ProQuest Platform.},
langid = {english},
keywords = {DFT,dissertation,ML,preview},
- file = {/home/johannes/Zotero/storage/WLWNEY4Q/1.html}
+ file = {/Users/wasmer/Zotero/storage/WLWNEY4Q/1.html}
}
@article{jablonkaBigDataSciencePorous2020,
@@ -3653,7 +3989,7 @@
urldate = {2021-05-30},
abstract = {By combining metal nodes with organic linkers we can potentially synthesize millions of possible metal–organic frameworks (MOFs). The fact that we have so many materials opens many exciting avenues but also create new challenges. We simply have too many materials to be processed using conventional, brute force, methods. In this review, we show that having so many materials allows us to use big-data methods as a powerful technique to study these materials and to discover complex correlations. The first part of the review gives an introduction to the principles of big-data science. We show how to select appropriate training sets, survey approaches that are used to represent these materials in feature space, and review different learning architectures, as well as evaluation and interpretation strategies. In the second part, we review how the different approaches of machine learning have been applied to porous materials. In particular, we discuss applications in the field of gas storage and separation, the stability of these materials, their electronic properties, and their synthesis. Given the increasing interest of the scientific community in machine learning, we expect this list to rapidly expand in the coming years.},
keywords = {descriptors,materials discovery,ML,MLP,review},
- file = {/home/johannes/Nextcloud/Zotero/Jablonka et al_2020_Big-Data Science in Porous Materials.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Jablonka et al_2020_Big-Data Science in Porous Materials.pdf}
}
@article{jacobsMaterialsSimulationToolkit2020,
@@ -3672,7 +4008,7 @@
abstract = {As data science and machine learning methods are taking on an increasingly important role in the materials research community, there is a need for the development of machine learning software tools that are easy to use (even for nonexperts with no programming ability), provide flexible access to the most important algorithms, and codify best practices of machine learning model development and evaluation. Here, we introduce the Materials Simulation Toolkit for Machine Learning (MAST-ML), an open source Python-based software package designed to broaden and accelerate the use of machine learning in materials science research. MAST-ML provides predefined routines for many input setup, model fitting, and post-analysis tasks, as well as a simple structure for executing a multi-step machine learning model workflow. In this paper, we describe how MAST-ML is used to streamline and accelerate the execution of machine learning problems. We walk through how to acquire and run MAST-ML, demonstrate how to execute different components of a supervised machine learning workflow via a customized input file, and showcase a number of features and analyses conducted automatically during a MAST-ML run. Further, we demonstrate the utility of MAST-ML by showcasing examples of recent materials informatics studies which used MAST-ML to formulate and evaluate various machine learning models for an array of materials applications. Finally, we lay out a vision of how MAST-ML, together with complementary software packages and emerging cyberinfrastructure, can advance the rapidly growing field of materials informatics, with a focus on producing machine learning models easily, reproducibly, and in a manner that facilitates model evolution and improvement in the future.},
langid = {english},
keywords = {library,materials informatics,with-code},
- file = {/home/johannes/Nextcloud/Zotero/Jacobs et al_2020_The Materials Simulation Toolkit for Machine learning (MAST-ML).pdf;/home/johannes/Zotero/storage/5R5YHQE4/S0927025620300355.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Jacobs et al_2020_The Materials Simulation Toolkit for Machine learning (MAST-ML).pdf;/Users/wasmer/Zotero/storage/5R5YHQE4/S0927025620300355.html}
}
@article{jainCommentaryMaterialsProject2013,
@@ -3690,7 +4026,7 @@
urldate = {2021-10-15},
abstract = {Accelerating the discovery of advanced materials is essential for human welfare and sustainable, clean energy. In this paper, we introduce the Materials Project (www.materialsproject.org), a core program of the Materials Genome Initiative that uses high-throughput computing to uncover the properties of all known inorganic materials. This open dataset can be accessed through multiple channels for both interactive exploration and data mining. The Materials Project also seeks to create open-source platforms for developing robust, sophisticated materials analyses. Future efforts will enable users to perform ‘‘rapid-prototyping’’ of new materials in silico, and provide researchers with new avenues for cost-effective, data-driven materials design.},
keywords = {materials project},
- file = {/home/johannes/Nextcloud/Zotero/Jain et al_2013_Commentary.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Jain et al_2013_Commentary.pdf}
}
@article{jainFireWorksDynamicWorkflow2015,
@@ -3709,8 +4045,7 @@
abstract = {This paper introduces FireWorks, a workflow software for running high-throughput calculation workflows at supercomputing centers. FireWorks has been used to complete over 50 million CPU-hours worth of computational chemistry and materials science calculations at the National Energy Research Supercomputing Center. It has been designed to serve the demanding high-throughput computing needs of these applications, with extensive support for (i) concurrent execution through job packing, (ii) failure detection and correction, (iii) provenance and reporting for long-running projects, (iv) automated duplicate detection, and (v) dynamic workflows (i.e., modifying the workflow graph during runtime). We have found that these features are highly relevant to enabling modern data-driven and high-throughput science applications, and we discuss our implementation strategy that rests on Python and NoSQL databases (MongoDB). Finally, we present performance data and limitations of our approach along with planned future work. Copyright © 2015 John Wiley \& Sons, Ltd.},
langid = {english},
keywords = {fault-tolerant computing,high-throughput computing,scientific workflows},
- annotation = {\_eprint: https://onlinelibrary.wiley.com/doi/pdf/10.1002/cpe.3505},
- file = {/home/johannes/Nextcloud/Zotero/Jain et al_2015_FireWorks.pdf;/home/johannes/Zotero/storage/FFWIWLTR/cpe.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Jain et al_2015_FireWorks.pdf;/Users/wasmer/Zotero/storage/FFWIWLTR/cpe.html}
}
@article{janssenPyironIntegratedDevelopment2019,
@@ -3729,7 +4064,7 @@
abstract = {To support and accelerate the development of simulation protocols in atomistic modelling, we introduce an integrated development environment (IDE) for computational materials science called pyiron (http://pyiron.org). The pyiron IDE combines a web based source code editor, a job management system for build automation, and a hierarchical data management solution. The core components of the pyiron IDE are pyiron objects based on an abstract class, which links application structures such as atomistic structures, projects, jobs, simulation protocols and computing resources with persistent storage and an interactive user environment. The simulation protocols within the pyiron IDE are constructed using the Python programming language. To highlight key concepts of this tool as well as to demonstrate its ability to simplify the implementation and testing of simulation protocols we discuss two applications. In these examples we show how pyiron supports the whole life cycle of a typical simulation, seamlessly combines ab initio with empirical potential calculations, and how complex feedback loops can be implemented. While originally developed with focus on ab initio thermodynamics simulations, the concepts and implementation of pyiron are general thus allowing to employ it for a wide range of simulation topics.},
langid = {english},
keywords = {Complex simulation protocols,Integrated development environment,Modelling workflow},
- file = {/home/johannes/Zotero/storage/TNV7XY35/S0927025618304786.html}
+ file = {/Users/wasmer/Zotero/storage/TNV7XY35/S0927025618304786.html}
}
@inproceedings{jiaPushingLimitMolecular2020,
@@ -3742,7 +4077,27 @@
abstract = {For 35 years, ab initio molecular dynamics (AIMD) has been the method of choice for modeling complex atomistic phenomena from first principles. However, most AIMD applications are limited by computational cost to systems with thousands of atoms at most. We report that a machine learning based simulation protocol (Deep Potential Molecular Dynamics), while retaining ab initio accuracy, can simulate more than 1 nanosecond-long trajectory of over 100 million atoms per day, using a highly optimized code (GPU DeePMD-kit) on the Summit supercomputer. Our code can efficiently scale up to the entire Summit supercomputer, attaining 91 PFLOPS in double precision (45.5\% of the peak) and 162/275 PFLOPS in mixed-single/half precision. The great accomplishment of this work is that it opens the door to simulating unprecedented size and time scales with ab initio accuracy. It also poses new challenges to the next-generation supercomputer for a better integration of machine learning and physical modeling.},
eventtitle = {{{SC20}}: {{International Conference}} for {{High Performance Computing}}, {{Networking}}, {{Storage}} and {{Analysis}}},
keywords = {100 million atoms,DeePMD-kit,MD,ML,MLP,record,Supercomputer},
- file = {/home/johannes/Nextcloud/Zotero/Jia et al_2020_Pushing the Limit of Molecular Dynamics with Ab Initio Accuracy to 100 Million.pdf;/home/johannes/Zotero/storage/UML425XW/9355242.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Jia et al_2020_Pushing the Limit of Molecular Dynamics with Ab Initio Accuracy to 100 Million.pdf;/Users/wasmer/Zotero/storage/UML425XW/9355242.html}
+}
+
+@article{jinHopesRaisedRoomtemperature2023,
+ title = {Hopes Raised for Room-Temperature Superconductivity, but Doubts Remain},
+ author = {Jin, ChangQing and Ceperley, David},
+ date = {2023-03},
+ journaltitle = {Nature},
+ volume = {615},
+ number = {7951},
+ pages = {221--222},
+ publisher = {{Nature Publishing Group}},
+ doi = {10.1038/d41586-023-00599-9},
+ url = {https://www.nature.com/articles/d41586-023-00599-9},
+ urldate = {2023-03-10},
+ abstract = {Evidence of a superconducting state at 294 kelvin.},
+ issue = {7951},
+ langid = {english},
+ keywords = {/unread,superconductor},
+ annotation = {Bandiera\_abtest: a Cg\_type: News And Views Subject\_term: Condensed-matter physics, Physics, Materials science},
+ file = {/Users/wasmer/Nextcloud/Zotero/Jin_Ceperley_2023_Hopes raised for room-temperature superconductivity, but doubts remain.pdf}
}
@book{johanssonNumericalPythonScientific2019,
@@ -3775,44 +4130,40 @@
langid = {english},
keywords = {FZJ,KKR,NIC,NIC winter school,PGI-1/IAS-1},
annotation = {OCLC: 77518371},
- file = {/home/johannes/Nextcloud/Zotero/John von Neumann-Institut für Computing et al_2006_Computational nanoscience.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/John von Neumann-Institut für Computing et al_2006_Computational nanoscience.pdf}
}
-@misc{jorgensenDeepDFTNeuralMessage2020,
+@online{jorgensenDeepDFTNeuralMessage2020,
title = {{{DeepDFT}}: {{Neural Message Passing Network}} for {{Accurate Charge Density Prediction}}},
shorttitle = {{{DeepDFT}}},
author = {Jørgensen, Peter Bjørn and Bhowmik, Arghya},
date = {2020-11-04},
number = {arXiv:2011.03346},
- eprint = {2011.03346},
+ eprint = {arXiv:2011.03346},
eprinttype = {arxiv},
- primaryclass = {physics},
- publisher = {{arXiv}},
doi = {10.48550/arXiv.2011.03346},
url = {http://arxiv.org/abs/2011.03346},
urldate = {2022-07-10},
abstract = {We introduce DeepDFT, a deep learning model for predicting the electronic charge density around atoms, the fundamental variable in electronic structure simulations from which all ground state properties can be calculated. The model is formulated as neural message passing on a graph, consisting of interacting atom vertices and special query point vertices for which the charge density is predicted. The accuracy and scalability of the model are demonstrated for molecules, solids and liquids. The trained model achieves lower average prediction errors than the observed variations in charge density obtained from density functional theory simulations using different exchange correlation functionals.},
- archiveprefix = {arXiv},
+ pubstate = {preprint},
keywords = {Computer Science - Machine Learning,Physics - Computational Physics},
- file = {/home/johannes/Nextcloud/Zotero/Jørgensen_Bhowmik_2020_DeepDFT.pdf;/home/johannes/Zotero/storage/QXJKV745/2011.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Jørgensen_Bhowmik_2020_DeepDFT.pdf;/Users/wasmer/Zotero/storage/QXJKV745/2011.html}
}
-@misc{jorgensenGraphNeuralNetworks2021,
+@online{jorgensenGraphNeuralNetworks2021,
title = {Graph Neural Networks for Fast Electron Density Estimation of Molecules, Liquids, and Solids},
author = {Jørgensen, Peter Bjørn and Bhowmik, Arghya},
date = {2021-12-01},
number = {arXiv:2112.00652},
- eprint = {2112.00652},
+ eprint = {arXiv:2112.00652},
eprinttype = {arxiv},
- primaryclass = {physics, stat},
- publisher = {{arXiv}},
doi = {10.48550/arXiv.2112.00652},
url = {http://arxiv.org/abs/2112.00652},
urldate = {2022-07-10},
abstract = {Electron density \$\textbackslash rho(\textbackslash vec\{r\})\$ is the fundamental variable in the calculation of ground state energy with density functional theory (DFT). Beyond total energy, features in \$\textbackslash rho(\textbackslash vec\{r\})\$ distribution and modifications in \$\textbackslash rho(\textbackslash vec\{r\})\$ are often used to capture critical physicochemical phenomena in functional materials and molecules at the electronic scale. Methods providing access to \$\textbackslash rho(\textbackslash vec\{r\})\$ of complex disordered systems with little computational cost can be a game changer in the expedited exploration of materials phase space towards the inverse design of new materials with better functionalities. We present a machine learning framework for the prediction of \$\textbackslash rho(\textbackslash vec\{r\})\$. The model is based on equivariant graph neural networks and the electron density is predicted at special query point vertices that are part of the message passing graph, but only receive messages. The model is tested across multiple data sets of molecules (QM9), liquid ethylene carbonate electrolyte (EC) and LixNiyMnzCo(1-y-z)O2 lithium ion battery cathodes (NMC). For QM9 molecules, the accuracy of the proposed model exceeds typical variability in \$\textbackslash rho(\textbackslash vec\{r\})\$ obtained from DFT done with different exchange-correlation functional and show beyond the state of the art accuracy. The accuracy is even better for the mixed oxide (NMC) and electrolyte (EC) datasets. The linear scaling model's capacity to probe thousands of points simultaneously permits calculation of \$\textbackslash rho(\textbackslash vec\{r\})\$ for large complex systems many orders of magnitude faster than DFT allowing screening of disordered functional materials.},
- archiveprefix = {arXiv},
+ pubstate = {preprint},
keywords = {Physics - Computational Physics,Statistics - Machine Learning},
- file = {/home/johannes/Nextcloud/Zotero/Jørgensen_Bhowmik_2021_Graph neural networks for fast electron density estimation of molecules,.pdf;/home/johannes/Zotero/storage/MBXG22TT/2112.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Jørgensen_Bhowmik_2021_Graph neural networks for fast electron density estimation of molecules,.pdf;/Users/wasmer/Zotero/storage/MBXG22TT/2112.html}
}
@online{julichJulichKKRCodes,
@@ -3823,7 +4174,28 @@
urldate = {2021-10-17},
abstract = {The Jülich family of codes implementing the Korringa-Kohn-Rostoker Green Function method for electronic structure calculations},
langid = {american},
- file = {/home/johannes/Zotero/storage/ZX3TRV7K/jukkr.fz-juelich.de.html}
+ file = {/Users/wasmer/Zotero/storage/ZX3TRV7K/jukkr.fz-juelich.de.html}
+}
+
+@article{kajitaDiscoverySuperionicConductors2020,
+ title = {Discovery of Superionic Conductors by Ensemble-Scope Descriptor},
+ author = {Kajita, Seiji and Ohba, Nobuko and Suzumura, Akitoshi and Tajima, Shin and Asahi, Ryoji},
+ date = {2020-04-24},
+ journaltitle = {NPG Asia Materials},
+ shortjournal = {NPG Asia Mater},
+ volume = {12},
+ number = {1},
+ pages = {1--8},
+ publisher = {{Nature Publishing Group}},
+ issn = {1884-4057},
+ doi = {10.1038/s41427-020-0211-1},
+ url = {https://www.nature.com/articles/s41427-020-0211-1},
+ urldate = {2023-03-12},
+ abstract = {Machine learning accelerates virtual screening in which material candidates are selected from existing databases, facilitating materials discovery in a broad chemical search space. Machine learning models quickly predict a target property from explanatory material features called descriptors. However, a major bottleneck of the machine learning model is an insufficient amount of training data in materials science, especially data with non-equilibrium properties. Here, we develop an alternative virtual-screening process via ensemble-based machine learning with one handcrafted and two generic descriptors to maximize the inference ability even using a small training dataset. A joint representation with the three descriptors translates the physical and chemical properties of a material as well as its underlying short- and long-range atomic structures to describe a multifaceted perspective of the material. As an application, the ensemble-scope descriptor learning model was trained with only 29 entries in the training dataset, and it selected potential oxygen-ion conductors from 13,384 oxides in the inorganic crystal structure database. The experiments confirmed that we successfully discovered five compounds that have not been reported, to the best of our knowledge, as oxygen-ion conductors.},
+ issue = {1},
+ langid = {english},
+ keywords = {/unread,CNN,compositional descriptors,defects,descriptor comparison,disordered,doping,KRR,model comparison,NN,prediction of ion conductivity,SOAP,supercell,VASP},
+ file = {/Users/wasmer/Nextcloud/Zotero/Kajita et al_2020_Discovery of superionic conductors by ensemble-scope descriptor.pdf}
}
@article{kalitaHowWellDoes2022,
@@ -3841,7 +4213,7 @@
urldate = {2022-07-10},
abstract = {Kohn–Sham regularizer (KSR) is a differentiable machine learning approach to finding the exchange-correlation functional in Kohn–Sham density functional theory that works for strongly correlated systems. Here we test KSR for a weak correlation. We propose spin-adapted KSR (sKSR) with trainable local, semilocal, and nonlocal approximations found by minimizing density and total energy loss. We assess the atoms-to-molecules generalizability by training on one-dimensional (1D) H, He, Li, Be, and Be2+ and testing on 1D hydrogen chains, LiH, BeH2, and helium hydride complexes. The generalization error from our semilocal approximation is comparable to other differentiable approaches, but our nonlocal functional outperforms any existing machine learning functionals, predicting ground-state energies of test systems with a mean absolute error of 2.7 mH.},
keywords = {DFT,Kohn-Sham regularizer,ML,ML-DFA,ML-DFT,ML-ESM,spin-dependent,spin-polarized},
- file = {/home/johannes/Nextcloud/Zotero/Kalita et al_2022_How Well Does Kohn–Sham Regularizer Work for Weakly Correlated Systems.pdf;/home/johannes/Zotero/storage/TCWGCAZA/acs.jpclett.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Kalita et al_2022_How Well Does Kohn–Sham Regularizer Work for Weakly Correlated Systems.pdf;/Users/wasmer/Zotero/storage/TCWGCAZA/acs.jpclett.html}
}
@article{kalitaLearningApproximateDensity2021,
@@ -3860,7 +4232,7 @@
urldate = {2021-12-14},
abstract = {ConspectusDensity functional theory (DFT) calculations are used in over 40,000 scientific papers each year, in chemistry, materials science, and far beyond. DFT is extremely useful because it is computationally much less expensive than ab initio electronic structure methods and allows systems of considerably larger size to be treated. However, the accuracy of any Kohn–Sham DFT calculation is limited by the approximation chosen for the exchange-correlation (XC) energy. For more than half a century, humans have developed the art of such approximations, using general principles, empirical data, or a combination of both, typically yielding useful results, but with errors well above the chemical accuracy limit (1 kcal/mol). Over the last 15 years, machine learning (ML) has made major breakthroughs in many applications and is now being applied to electronic structure calculations. This recent rise of ML begs the question: Can ML propose or improve density functional approximations? Success could greatly enhance the accuracy and usefulness of DFT calculations without increasing the cost.In this work, we detail efforts in this direction, beginning with an elementary proof of principle from 2012, namely, finding the kinetic energy of several Fermions in a box using kernel ridge regression. This is an example of orbital-free DFT, for which a successful general-purpose scheme could make even DFT calculations run much faster. We trace the development of that work to state-of-the-art molecular dynamics simulations of resorcinol with chemical accuracy. By training on ab initio examples, one bypasses the need to find the XC functional explicitly. We also discuss how the exchange-correlation energy itself can be modeled with such methods, especially for strongly correlated materials. Finally, we show how deep neural networks with differentiable programming can be used to construct accurate density functionals from very few data points by using the Kohn–Sham equations themselves as a regularizer. All these cases show that ML can create approximations of greater accuracy than humans, and is capable of finding approximations that can deal with difficult cases such as strong correlation. However, such ML-designed functionals have not been implemented in standard codes because of one last great challenge: generalization. We discuss how effortlessly human-designed functionals can be applied to a wide range of situations, and how difficult that is for ML.},
keywords = {DFT,Kohn-Sham regularizer,ML,ML-DFA,ML-DFT,ML-ESM,OF-DFT,review},
- file = {/home/johannes/Nextcloud/Zotero/Kalita et al_2021_Learning to Approximate Density Functionals.pdf;/home/johannes/Zotero/storage/MUYRWPH9/acs.accounts.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Kalita et al_2021_Learning to Approximate Density Functionals.pdf;/Users/wasmer/Zotero/storage/MUYRWPH9/acs.accounts.html}
}
@unpublished{kalitaUsingMachineLearning2021,
@@ -3869,13 +4241,12 @@
date = {2021-12-03},
eprint = {2112.05554},
eprinttype = {arxiv},
- primaryclass = {physics},
+ eprintclass = {physics},
url = {http://arxiv.org/abs/2112.05554},
urldate = {2022-03-28},
abstract = {Machine learning has now become an integral part of research and innovation. The field of machine learning density functional theory has continuously expanded over the years while making several noticeable advances. We briefly discuss the status of this field and point out some current and future challenges. We also talk about how state-of-the-art science and technology tools can help overcome these challenges. This draft is a part of the "Roadmap on Machine Learning in Electronic Structure" to be published in Electronic Structure (EST).},
- archiveprefix = {arXiv},
keywords = {DFT,Kohn-Sham regularizer,ML,ML-DFA,ML-DFT,ML-ESM},
- file = {/home/johannes/Nextcloud/Zotero/Kalita_Burke_2021_Using Machine Learning to Find New Density Functionals.pdf;/home/johannes/Zotero/storage/6FMA3TRD/2112.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Kalita_Burke_2021_Using Machine Learning to Find New Density Functionals.pdf;/Users/wasmer/Zotero/storage/6FMA3TRD/2112.html}
}
@inproceedings{kanterDeepFeatureSynthesis2015,
@@ -3889,7 +4260,7 @@
abstract = {In this paper, we develop the Data Science Machine, which is able to derive predictive models from raw data automatically. To achieve this automation, we first propose and develop the Deep Feature Synthesis algorithm for automatically generating features for relational datasets. The algorithm follows relationships in the data to a base field, and then sequentially applies mathematical functions along that path to create the final feature. Second, we implement a generalizable machine learning pipeline and tune it using a novel Gaussian Copula process based approach. We entered the Data Science Machine in 3 data science competitions that featured 906 other data science teams. Our approach beats 615 teams in these data science competitions. In 2 of the 3 competitions we beat a majority of competitors, and in the third, we achieved 94\% of the best competitor's score. In the best case, with an ongoing competition, we beat 85.6\% of the teams and achieved 95.7\% of the top submissions score.},
eventtitle = {2015 {{IEEE International Conference}} on {{Data Science}} and {{Advanced Analytics}} ({{DSAA}})},
keywords = {Data models,Feature extraction,ML,thesis},
- file = {/home/johannes/Nextcloud/Zotero/Kanter_Veeramachaneni_2015_Deep feature synthesis.pdf;/home/johannes/Zotero/storage/FTYFE5ZI/7344858.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Kanter_Veeramachaneni_2015_Deep feature synthesis.pdf;/Users/wasmer/Zotero/storage/FTYFE5ZI/7344858.html}
}
@article{kasimLearningExchangecorrelationFunctional2021,
@@ -3908,9 +4279,8 @@
url = {http://arxiv.org/abs/2102.04229},
urldate = {2022-01-02},
abstract = {Improving the predictive capability of molecular properties in ab initio simulations is essential for advanced material discovery. Despite recent progress making use of machine learning, utilizing deep neural networks to improve quantum chemistry modelling remains severely limited by the scarcity and heterogeneity of appropriate experimental data. Here we show how training a neural network to replace the exchange-correlation functional within a fully-differentiable three-dimensional Kohn-Sham density functional theory (DFT) framework can greatly improve simulation accuracy. Using only eight experimental data points on diatomic molecules, our trained exchange-correlation networks enable improved prediction accuracy of atomization energies across a collection of 104 molecules containing new bonds and atoms that are not present in the training dataset.},
- archiveprefix = {arXiv},
keywords = {Computer Science - Machine Learning,Physics - Chemical Physics,Physics - Computational Physics},
- file = {/home/johannes/Nextcloud/Zotero/Kasim_Vinko_2021_Learning the exchange-correlation functional from nature with fully.pdf;/home/johannes/Zotero/storage/TFQLR3CJ/2102.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Kasim_Vinko_2021_Learning the exchange-correlation functional from nature with fully.pdf;/Users/wasmer/Zotero/storage/TFQLR3CJ/2102.html}
}
@article{kaundinyaPredictionElectronDensity2022,
@@ -3929,7 +4299,7 @@
abstract = {Machine learning (ML)-based models have greatly enhanced the traditional materials discovery and design pipeline. Specifically, in recent years, surrogate ML models for material property prediction have demonstrated success in predicting discrete scalar-valued target properties to within reasonable accuracy of their DFT-computed values. However, accurate prediction of spectral targets, such as the electron density of states (DOS), poses a much more challenging problem due to the complexity of the target, and the limited amount of available training data. In this study, we present an extension of the recently developed atomistic line graph neural network to accurately predict DOS of a large set of material unit cell structures, trained to the publicly available JARVIS-DFT dataset. Furthermore, we evaluate two methods of representation of the target quantity: a direct discretized spectrum, and a compressed low-dimensional representation obtained using an autoencoder. Through this work, we demonstrate the utility of graph-based featurization and modeling methods in the prediction of complex targets that depend on both chemistry and directional characteristics of material structures.},
langid = {english},
keywords = {_tablet,ALIGNN,autoencoder,DFT,dimensionality reduction,dimensionality reduction of target,GNN,JARVIS,JARVIS-DFT,ML,ML-DFT,ML-ESM,MPNN,prediction from structure,prediction of LDOS},
- file = {/home/johannes/Nextcloud/Zotero/Kaundinya et al_2022_Prediction of the Electron Density of States for Crystalline Compounds with.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Kaundinya et al_2022_Prediction of the Electron Density of States for Crystalline Compounds with.pdf}
}
@article{keimerPhysicsQuantumMaterials2017,
@@ -3950,7 +4320,7 @@
issue = {11},
langid = {english},
annotation = {Bandiera\_abtest: a Cg\_type: Nature Research Journals Primary\_atype: Reviews Subject\_term: Quantum physics;Theoretical physics Subject\_term\_id: quantum-physics;theoretical-physics},
- file = {/home/johannes/Nextcloud/Zotero/Keimer_Moore_2017_The physics of quantum materials.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Keimer_Moore_2017_The physics of quantum materials.pdf}
}
@article{khorshidiAmpModularApproach2016,
@@ -3969,25 +4339,23 @@
abstract = {Electronic structure calculations, such as those employing Kohn–Sham density functional theory or ab initio wavefunction theories, have allowed for atomistic-level understandings of a wide variety of phenomena and properties of matter at small scales. However, the computational cost of electronic structure methods drastically increases with length and time scales, which makes these methods difficult for long time-scale molecular dynamics simulations or large-sized systems. Machine-learning techniques can provide accurate potentials that can match the quality of electronic structure calculations, provided sufficient training data. These potentials can then be used to rapidly simulate large and long time-scale phenomena at similar quality to the parent electronic structure approach. Machine-learning potentials usually take a bias-free mathematical form and can be readily developed for a wide variety of systems. Electronic structure calculations have favorable properties–namely that they are noiseless and targeted training data can be produced on-demand–that make them particularly well-suited for machine learning. This paper discusses our modular approach to atomistic machine learning through the development of the open-source Atomistic Machine-learning Package (Amp), which allows for representations of both the total and atom-centered potential energy surface, in both periodic and non-periodic systems. Potentials developed through the atom-centered approach are simultaneously applicable for systems with various sizes. Interpolation can be enhanced by introducing custom descriptors of the local environment. We demonstrate this in the current work for Gaussian-type, bispectrum, and Zernike-type descriptors. Amp ~has an intuitive and modular structure with an interface through the python scripting language yet has parallelizable fortran components for demanding tasks; it is designed to integrate closely with the widely used Atomic Simulation Environment (ASE), which makes it compatible with a wide variety of commercial and open-source electronic structure codes. We finally demonstrate that the neural network model inside Amp ~can accurately interpolate electronic structure energies as well as forces of thousands of multi-species atomic systems. Program summary Program title: Amp Catalogue identifier: AFAK\_v1\_0 Program summary URL:http://cpc.cs.qub.ac.uk/summaries/AFAK\_v1\_0.html Program obtainable from: CPC Program Library, Queen’s University, Belfast, N. Ireland Licensing provisions: yes No. of lines in distributed program, including test data, etc.: 21239 No. of bytes in distributed program, including test data, etc.: 1412975 Distribution format: tar.gz Programming language: Python, Fortran. Computer: PC, Mac. Operating system: Linux, Mac, Windows. Has the code been vectorized or parallelized?: Yes RAM: Variable, depending on the number and size of atomic systems. Classification: 16.1, 2.1. External routines: ASE, NumPy, SciPy, f2py, matplotlib Nature of problem: Atomic interactions within many-body systems typically have complicated functional forms, difficult to represent in simple pre-decided closed-forms. Solution method: Machine learning provides flexible functional forms that can be improved as new situations are encountered. Typically, interatomic potentials yield from machine learning simultaneously apply to different system sizes. Unusual features: Amp is as modular as possible, providing a framework for the user to create atomic environment descriptor and regression model at will. Moreover, it has Atomic Simulation Environment (ASE) interface, facilitating interactive collaboration with other electronic structure calculators within ASE. Running time: Variable, depending on the number and size of atomic systems.},
langid = {english},
keywords = {ACSF,ase,bispectrum,BPNN,BPSF,descriptors,DFT,library,MD,ML,MLP,models,neural network potentials,PES,SOAP,with-code,Zernike descriptors},
- file = {/home/johannes/Nextcloud/Zotero/Khorshidi_Peterson_2016_Amp.pdf;/home/johannes/Zotero/storage/PFLW4RH4/S0010465516301266.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Khorshidi_Peterson_2016_Amp.pdf;/Users/wasmer/Zotero/storage/PFLW4RH4/S0010465516301266.html}
}
-@misc{kidgerNeuralDifferentialEquations2022,
+@online{kidgerNeuralDifferentialEquations2022,
title = {On {{Neural Differential Equations}}},
author = {Kidger, Patrick},
date = {2022-02-04},
number = {arXiv:2202.02435},
- eprint = {2202.02435},
+ eprint = {arXiv:2202.02435},
eprinttype = {arxiv},
- primaryclass = {cs, math, stat},
- publisher = {{arXiv}},
doi = {10.48550/arXiv.2202.02435},
url = {http://arxiv.org/abs/2202.02435},
urldate = {2022-09-07},
abstract = {The conjoining of dynamical systems and deep learning has become a topic of great interest. In particular, neural differential equations (NDEs) demonstrate that neural networks and differential equation are two sides of the same coin. Traditional parameterised differential equations are a special case. Many popular neural network architectures, such as residual networks and recurrent networks, are discretisations. NDEs are suitable for tackling generative problems, dynamical systems, and time series (particularly in physics, finance, ...) and are thus of interest to both modern machine learning and traditional mathematical modelling. NDEs offer high-capacity function approximation, strong priors on model space, the ability to handle irregular data, memory efficiency, and a wealth of available theory on both sides. This doctoral thesis provides an in-depth survey of the field. Topics include: neural ordinary differential equations (e.g. for hybrid neural/mechanistic modelling of physical systems); neural controlled differential equations (e.g. for learning functions of irregular time series); and neural stochastic differential equations (e.g. to produce generative models capable of representing complex stochastic dynamics, or sampling from complex high-dimensional distributions). Further topics include: numerical methods for NDEs (e.g. reversible differential equations solvers, backpropagation through differential equations, Brownian reconstruction); symbolic regression for dynamical systems (e.g. via regularised evolution); and deep implicit models (e.g. deep equilibrium models, differentiable optimisation). We anticipate this thesis will be of interest to anyone interested in the marriage of deep learning with dynamical systems, and hope it will provide a useful reference for the current state of the art.},
- archiveprefix = {arXiv},
+ pubstate = {preprint},
keywords = {backpropagation,Deep learning,differential equations,NDE,PINN,thesis},
- file = {/home/johannes/Nextcloud/Zotero/Kidger_2022_On Neural Differential Equations.pdf;/home/johannes/Zotero/storage/EHARV7VZ/2202.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Kidger_2022_On Neural Differential Equations.pdf;/Users/wasmer/Zotero/storage/EHARV7VZ/2202.html}
}
@article{kippChiralHallEffect2021,
@@ -4008,7 +4376,7 @@
issue = {1},
langid = {english},
keywords = {Magnetic properties and materials,Spintronics},
- file = {/home/johannes/Nextcloud/Zotero/Kipp et al_2021_The chiral Hall effect in canted ferromagnets and antiferromagnets.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Kipp et al_2021_The chiral Hall effect in canted ferromagnets and antiferromagnets.pdf}
}
@article{kirkpatrickPushingFrontiersDensity2021,
@@ -4024,7 +4392,7 @@
url = {https://www.science.org/doi/10.1126/science.abj6511},
urldate = {2022-05-13},
keywords = {DeepMind,density functional,DFT,DM21,ML,ML-DFA,ML-DFT,ML-ESM,molecules,original publication,with-code},
- file = {/home/johannes/Nextcloud/Zotero/Kirkpatrick et al_2021_Pushing the frontiers of density functionals by solving the fractional electron.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Kirkpatrick et al_2021_Pushing the frontiers of density functionals by solving the fractional electron.pdf}
}
@unpublished{klicperaGemNetUniversalDirectional2021,
@@ -4034,13 +4402,12 @@
date = {2021-11-29},
eprint = {2106.08903},
eprinttype = {arxiv},
- primaryclass = {physics, stat},
+ eprintclass = {physics, stat},
url = {http://arxiv.org/abs/2106.08903},
urldate = {2022-01-02},
abstract = {Effectively predicting molecular interactions has the potential to accelerate molecular dynamics by multiple orders of magnitude and thus revolutionize chemical simulations. Graph neural networks (GNNs) have recently shown great successes for this task, overtaking classical methods based on fixed molecular kernels. However, they still appear very limited from a theoretical perspective, since regular GNNs cannot distinguish certain types of graphs. In this work we close this gap between theory and practice. We show that GNNs with directed edge embeddings and two-hop message passing are indeed universal approximators for predictions that are invariant to translation, and equivariant to permutation and rotation. We then leverage these insights and multiple structural improvements to propose the geometric message passing neural network (GemNet). We demonstrate the benefits of the proposed changes in multiple ablation studies. GemNet outperforms previous models on the COLL, MD17, and OC20 datasets by 34\%, 41\%, and 20\%, respectively, and performs especially well on the most challenging molecules. Our implementation is available online.},
- archiveprefix = {arXiv},
keywords = {Computer Science - Machine Learning,Physics - Chemical Physics,Physics - Computational Physics,Statistics - Machine Learning},
- file = {/home/johannes/Nextcloud/Zotero/Klicpera et al_2021_GemNet.pdf;/home/johannes/Zotero/storage/NPJ8A2J3/2106.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Klicpera et al_2021_GemNet.pdf;/Users/wasmer/Zotero/storage/NPJ8A2J3/2106.html}
}
@article{klintenbergComputationalSearchStrong2014,
@@ -4060,7 +4427,7 @@
issue = {4},
langid = {english},
keywords = {classification,DFT,FP-LMTO,LDA,materials screening,topological insulator},
- file = {/home/johannes/Nextcloud/Zotero/Klintenberg et al_2014_Computational Search for Strong Topological Insulators3.pdf;/home/johannes/Nextcloud/Zotero/Klintenberg et al_2014_Computational Search for Strong Topological Insulators4.pdf;/home/johannes/Zotero/storage/9MDA4BT7/37961.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Klintenberg et al_2014_Computational Search for Strong Topological Insulators3.pdf;/Users/wasmer/Nextcloud/Zotero/Klintenberg et al_2014_Computational Search for Strong Topological Insulators4.pdf;/Users/wasmer/Zotero/storage/9MDA4BT7/37961.html}
}
@unpublished{klusSymmetricAntisymmetricKernels2021,
@@ -4069,13 +4436,12 @@
date = {2021-03-31},
eprint = {2103.17233},
eprinttype = {arxiv},
- primaryclass = {math-ph, physics:physics, physics:quant-ph, stat},
+ eprintclass = {math-ph, physics:physics, physics:quant-ph, stat},
url = {http://arxiv.org/abs/2103.17233},
urldate = {2021-05-13},
abstract = {We derive symmetric and antisymmetric kernels by symmetrizing and antisymmetrizing conventional kernels and analyze their properties. In particular, we compute the feature space dimensions of the resulting polynomial kernels, prove that the reproducing kernel Hilbert spaces induced by symmetric and antisymmetric Gaussian kernels are dense in the space of symmetric and antisymmetric functions, and propose a Slater determinant representation of the antisymmetric Gaussian kernel, which allows for an efficient evaluation even if the state space is high-dimensional. Furthermore, we show that by exploiting symmetries or antisymmetries the size of the training data set can be significantly reduced. The results are illustrated with guiding examples and simple quantum physics and chemistry applications.},
- archiveprefix = {arXiv},
keywords = {kernel methods,ML,models},
- file = {/home/johannes/Nextcloud/Zotero/Klus et al_2021_Symmetric and antisymmetric kernels for machine learning problems in quantum.pdf;/home/johannes/Zotero/storage/WM8YDGB2/2103.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Klus et al_2021_Symmetric and antisymmetric kernels for machine learning problems in quantum.pdf;/Users/wasmer/Zotero/storage/WM8YDGB2/2103.html}
}
@unpublished{knosgaardRepresentingIndividualElectronic2021,
@@ -4084,13 +4450,12 @@
date = {2021-07-13},
eprint = {2107.06029},
eprinttype = {arxiv},
- primaryclass = {physics},
+ eprintclass = {physics},
url = {http://arxiv.org/abs/2107.06029},
urldate = {2021-08-05},
abstract = {We address the problem of representing quantum states of electrons in a solid for the purpose of machine leaning state-specific electronic properties. Specifically, we construct a fingerprint based on energy decomposed operator matrix elements (ENDOME) and radially decomposed projected density of states (RAD-PDOS), which are both obtainable from a standard density functional theory (DFT) calculation. Using such fingerprints we train a gradient boosting model on a set of 46k G\$\_0\$W\$\_0\$ quasiparticle energies. The resulting model predicts the self-energy correction of states in materials not seen by the model with a mean absolute error of 0.14 eV. By including the material's calculated dielectric constant in the fingerprint the error can be further reduced by 30\%, which we find is due to an enhanced ability to learn the correlation/screening part of the self-energy. Our work paves the way for accurate estimates of quasiparticle band structures at the cost of a standard DFT calculation.},
- archiveprefix = {arXiv},
keywords = {descriptors,DFT,electronic state,electronic state descriptors,ENDOME,materials,ML},
- file = {/home/johannes/Nextcloud/Zotero/Knøsgaard_Thygesen_2021_Representing individual electronic states in crystals for machine learning.pdf;/home/johannes/Zotero/storage/HWNEX2AE/2107.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Knøsgaard_Thygesen_2021_Representing individual electronic states in crystals for machine learning.pdf;/Users/wasmer/Zotero/storage/HWNEX2AE/2107.html}
}
@article{kocerContinuousOptimallyComplete2020,
@@ -4107,7 +4472,7 @@
url = {https://aip.scitation.org/doi/abs/10.1063/1.5111045},
urldate = {2021-05-13},
keywords = {descriptors,ML,MLP,SB descriptors,SOAP,Zernike descriptors},
- file = {/home/johannes/Nextcloud/Zotero/Kocer et al_2020_Continuous and optimally complete description of chemical environments using.pdf;/home/johannes/Zotero/storage/NZCCBRZE/1.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Kocer et al_2020_Continuous and optimally complete description of chemical environments using.pdf;/Users/wasmer/Zotero/storage/NZCCBRZE/1.html}
}
@article{koFourthgenerationHighdimensionalNeural2021,
@@ -4129,7 +4494,7 @@
langid = {english},
keywords = {CENT,HDNNP,rec-by-bluegel},
annotation = {Bandiera\_abtest: a Cc\_license\_type: cc\_by Cg\_type: Nature Research Journals Primary\_atype: Research Subject\_term: Computational methods;Density functional theory;Method development;Molecular dynamics Subject\_term\_id: computational-methods;density-functional-theory;method-development;molecular-dynamics},
- file = {/home/johannes/Nextcloud/Zotero/Ko et al_2021_A fourth-generation high-dimensional neural network potential with accurate.pdf;/home/johannes/Zotero/storage/2Z8H4HHW/s41467-020-20427-2.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Ko et al_2021_A fourth-generation high-dimensional neural network potential with accurate.pdf;/Users/wasmer/Zotero/storage/2Z8H4HHW/s41467-020-20427-2.html}
}
@article{koGeneralPurposeMachineLearning2021,
@@ -4148,7 +4513,7 @@
urldate = {2021-05-18},
abstract = {ConspectusThe development of first-principles-quality machine learning potentials (MLP) has seen tremendous progress, now enabling computer simulations of complex systems for which sufficiently accurate interatomic potentials have not been available. These advances and the increasing use of MLPs for more and more diverse systems gave rise to new questions regarding their applicability and limitations, which has constantly driven new developments. The resulting MLPs can be classified into several generations depending on the types of systems they are able to describe. First-generation MLPs, as introduced 25 years ago, have been applicable to low-dimensional systems such as small molecules. MLPs became a practical tool for complex systems in chemistry and materials science with the introduction of high-dimensional neural network potentials (HDNNP) in 2007, which represented the first MLP of the second generation. Second-generation MLPs are based on the concept of locality and express the total energy as a sum of environment-dependent atomic energies, which allows applications to very large systems containing thousands of atoms with linearly scaling computational costs. Since second-generation MLPs do not consider interactions beyond the local chemical environments, a natural extension has been the inclusion of long-range interactions without truncation, mainly electrostatics, employing environment-dependent charges establishing the third MLP generation. A variety of second- and, to some extent, also third-generation MLPs are currently the standard methods in ML-based atomistic simulations.In spite of countless successful applications, in recent years it has been recognized that the accuracy of MLPs relying on local atomic energies and charges is still insufficient for systems with long-ranged dependencies in the electronic structure. These can, for instance, result from nonlocal charge transfer or ionization and are omnipresent in many important types of systems and chemical processes such as the protonation and deprotonation of organic and biomolecules, redox reactions, and defects and doping in materials. In all of these situations, small local modifications can change the system globally, resulting in different equilibrium structures, charge distributions, and reactivity. These phenomena cannot be captured by second- and third-generation MLPs. Consequently, the inclusion of nonlocal phenomena has been identified as a next key step in the development of a new fourth generation of MLPs. While a first fourth-generation MLP, the charge equilibration neural network technique (CENT), was introduced in 2015, only very recently have a range of new general-purpose methods applicable to a broad range of physical scenarios emerged. In this Account, we show how fourth-generation HDNNPs can be obtained by combining the concepts of CENT and second-generation HDNNPs. These new MLPs allow for a highly accurate description of systems where nonlocal charge transfer is important.},
keywords = {HDNNP,long-range interaction,ML,MLP,models},
- file = {/home/johannes/Nextcloud/Zotero/Ko et al_2021_General-Purpose Machine Learning Potentials Capturing Nonlocal Charge Transfer.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Ko et al_2021_General-Purpose Machine Learning Potentials Capturing Nonlocal Charge Transfer.pdf}
}
@article{kohnSelfConsistentEquationsIncluding1965,
@@ -4162,7 +4527,7 @@
doi = {10.1103/PhysRev.140.A1133},
issue = {4A},
keywords = {DFT,KS-DFT,original publication},
- file = {/home/johannes/Nextcloud/Zotero/Kohn_1965_Self-Consistent Equations Including Exchange and Correlation Effects.pdf;/home/johannes/Zotero/storage/4CF9DCKS/PhysRev.140.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Kohn_1965_Self-Consistent Equations Including Exchange and Correlation Effects.pdf;/Users/wasmer/Zotero/storage/4CF9DCKS/PhysRev.140.html}
}
@article{korshunovaOpenChemDeepLearning2021,
@@ -4182,7 +4547,7 @@
urldate = {2021-07-21},
abstract = {Deep learning models have demonstrated outstanding results in many data-rich areas of research, such as computer vision and natural language processing. Currently, there is a rise of deep learning in computational chemistry and materials informatics, where deep learning could be effectively applied in modeling the relationship between chemical structures and their properties. With the immense growth of chemical and materials data, deep learning models can begin to outperform conventional machine learning techniques such as random forest, support vector machines, and nearest neighbor. Herein, we introduce OpenChem, a PyTorch-based deep learning toolkit for computational chemistry and drug design. OpenChem offers easy and fast model development, modular software design, and several data preprocessing modules. It is freely available via the GitHub repository.},
keywords = {chemistry,GCN,GNN,library,ML,models,pytorch},
- file = {/home/johannes/Nextcloud/Zotero/Korshunova et al_2021_OpenChem.pdf;/home/johannes/Zotero/storage/U5ZHRH93/acs.jcim.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Korshunova et al_2021_OpenChem.pdf;/Users/wasmer/Zotero/storage/U5ZHRH93/acs.jcim.html}
}
@article{kosmaStrongSpinorbitTorque2020,
@@ -4196,26 +4561,24 @@
number = {14},
doi = {10.1103/PhysRevB.102.144424},
keywords = {Boltzmann theory,DFT,IFF,KKR,PGI-1/IAS-1,rec-by-ruess,Spin-orbit effects,surface physics,topological insulator,topological spin textures,transport properties},
- file = {/home/johannes/Nextcloud/Zotero/Kosma_2020_Strong spin-orbit torque effect on magnetic defects due to topological surface.pdf;/home/johannes/Zotero/storage/5JLDY6FT/PhysRevB.102.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Kosma_2020_Strong spin-orbit torque effect on magnetic defects due to topological surface.pdf;/Users/wasmer/Zotero/storage/5JLDY6FT/PhysRevB.102.html}
}
-@misc{krennPredictingFutureAI2022,
+@online{krennPredictingFutureAI2022,
title = {Predicting the {{Future}} of {{AI}} with {{AI}}: {{High-quality}} Link Prediction in an Exponentially Growing Knowledge Network},
shorttitle = {Predicting the {{Future}} of {{AI}} with {{AI}}},
author = {Krenn, Mario and Buffoni, Lorenzo and Coutinho, Bruno and Eppel, Sagi and Foster, Jacob Gates and Gritsevskiy, Andrew and Lee, Harlin and Lu, Yichao and Moutinho, Joao P. and Sanjabi, Nima and Sonthalia, Rishi and Tran, Ngoc Mai and Valente, Francisco and Xie, Yangxinyu and Yu, Rose and Kopp, Michael},
date = {2022-09-23},
number = {arXiv:2210.00881},
- eprint = {2210.00881},
+ eprint = {arXiv:2210.00881},
eprinttype = {arxiv},
- primaryclass = {cs},
- publisher = {{arXiv}},
doi = {10.48550/arXiv.2210.00881},
url = {http://arxiv.org/abs/2210.00881},
urldate = {2022-10-05},
abstract = {A tool that could suggest new personalized research directions and ideas by taking insights from the scientific literature could significantly accelerate the progress of science. A field that might benefit from such an approach is artificial intelligence (AI) research, where the number of scientific publications has been growing exponentially over the last years, making it challenging for human researchers to keep track of the progress. Here, we use AI techniques to predict the future research directions of AI itself. We develop a new graph-based benchmark based on real-world data -- the Science4Cast benchmark, which aims to predict the future state of an evolving semantic network of AI. For that, we use more than 100,000 research papers and build up a knowledge network with more than 64,000 concept nodes. We then present ten diverse methods to tackle this task, ranging from pure statistical to pure learning methods. Surprisingly, the most powerful methods use a carefully curated set of network features, rather than an end-to-end AI approach. It indicates a great potential that can be unleashed for purely ML approaches without human knowledge. Ultimately, better predictions of new future research directions will be a crucial component of more advanced research suggestion tools.},
- archiveprefix = {arXiv},
+ pubstate = {preprint},
keywords = {General ML,literature analysis},
- file = {/home/johannes/Nextcloud/Zotero/Krenn et al_2022_Predicting the Future of AI with AI.pdf;/home/johannes/Zotero/storage/MZBX2N4K/2210.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Krenn et al_2022_Predicting the Future of AI with AI.pdf;/Users/wasmer/Zotero/storage/MZBX2N4K/2210.html}
}
@article{krennSelfreferencingEmbeddedStrings2020,
@@ -4236,7 +4599,7 @@
abstract = {The discovery of novel materials and functional molecules can help to solve some of society’s most urgent challenges, ranging from efficient energy harvesting and storage to uncovering novel pharmaceutical drug candidates. Traditionally matter engineering–generally denoted as inverse design–was based massively on human intuition and high-throughput virtual screening. The last few years have seen the emergence of significant interest in computer-inspired designs based on evolutionary or deep learning methods. The major challenge here is that the standard strings molecular representation SMILES shows substantial weaknesses in that task because large fractions of strings do not correspond to valid molecules. Here, we solve this problem at a fundamental level and introduce SELFIES (SELF-referencIng Embedded Strings), a string-based representation of molecules which is 100\% robust. Every SELFIES string corresponds to a valid molecule, and SELFIES can represent every molecule. SELFIES can be directly applied in arbitrary machine learning models without the adaptation of the models; each of the generated molecule candidates is valid. In our experiments, the model’s internal memory stores two orders of magnitude more diverse molecules than a similar test with SMILES. Furthermore, as all molecules are valid, it allows for explanation and interpretation of the internal working of the generative models.},
langid = {english},
keywords = {chemistry,descriptors,GAN,library,ML,molecules,SELFIES,SMILES,VAE},
- file = {/home/johannes/Nextcloud/Zotero/Krenn et al_2020_Self-referencing embedded strings (SELFIES).pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Krenn et al_2020_Self-referencing embedded strings (SELFIES).pdf}
}
@article{kulikRoadmapMachineLearning2022,
@@ -4252,7 +4615,7 @@
abstract = {In recent years, we have been witnessing a paradigm shift in computational materials science. In fact, traditional methods, mostly developed in the second half of the XXth century, are being complemented, extended, and sometimes even completely replaced by faster, simpler, and often more accurate approaches. The new approaches, that we collectively label by machine learning, have their origins in the fields of informatics and artificial intelligence, but are making rapid inroads in all other branches of science. With this in mind, this Roadmap article, consisting of multiple contributions from experts across the field, discusses the use of machine learning in materials science, and share perspectives on current and future challenges in problems as diverse as the prediction of materials properties, the construction of force-fields, the development of exchange correlation functionals for density-functional theory, the solution of the many-body problem, and more. In spite of the already numerous and exciting success stories, we are just at the beginning of a long path that will reshape materials science for the many challenges of the XXIth century.},
langid = {english},
keywords = {_tablet,descriptors,DFT,electronic structure theory,MD,ML,ML-DFT,ML-ESM,models,review,roadmap,surrogate model},
- file = {/home/johannes/Nextcloud/Zotero/false;/home/johannes/Nextcloud/Zotero/Kulik et al_2022_Roadmap on Machine Learning in Electronic Structure.pdf}
+ file = {/home/johannes/Nextcloud/Zotero/false;/Users/wasmer/Nextcloud/Zotero/Kulik et al_2022_Roadmap on Machine Learning in Electronic Structure.pdf}
}
@article{kumarTopologicalQuantumMaterials2021,
@@ -4271,26 +4634,24 @@
urldate = {2021-05-13},
abstract = {Topology, a mathematical concept, has recently become a popular and truly transdisciplinary topic encompassing condensed matter physics, solid state chemistry, and materials science. Since there is a direct connection between real space, namely atoms, valence electrons, bonds, and orbitals, and reciprocal space, namely bands and Fermi surfaces, via symmetry and topology, classifying topological materials within a single-particle picture is possible. Currently, most materials are classified as trivial insulators, semimetals, and metals or as topological insulators, Dirac and Weyl nodal-line semimetals, and topological metals. The key ingredients for topology are certain symmetries, the inert pair effect of the outer electrons leading to inversion of the conduction and valence bands, and spin–orbit coupling. This review presents the topological concepts related to solids from the viewpoint of a solid-state chemist, summarizes techniques for growing single crystals, and describes basic physical property measurement techniques to characterize topological materials beyond their structure and provide examples of such materials. Finally, a brief outlook on the impact of topology in other areas of chemistry is provided at the end of the article.},
keywords = {_tablet,chemistry,topological insulator},
- file = {/home/johannes/Nextcloud/Zotero/Kumar et al_2021_Topological Quantum Materials from the Viewpoint of Chemistry.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Kumar et al_2021_Topological Quantum Materials from the Viewpoint of Chemistry.pdf}
}
-@misc{lamGraphCastLearningSkillful2022,
+@online{lamGraphCastLearningSkillful2022,
title = {{{GraphCast}}: {{Learning}} Skillful Medium-Range Global Weather Forecasting},
shorttitle = {{{GraphCast}}},
author = {Lam, Remi and Sanchez-Gonzalez, Alvaro and Willson, Matthew and Wirnsberger, Peter and Fortunato, Meire and Pritzel, Alexander and Ravuri, Suman and Ewalds, Timo and Alet, Ferran and Eaton-Rosen, Zach and Hu, Weihua and Merose, Alexander and Hoyer, Stephan and Holland, George and Stott, Jacklynn and Vinyals, Oriol and Mohamed, Shakir and Battaglia, Peter},
date = {2022-12-24},
number = {arXiv:2212.12794},
- eprint = {2212.12794},
+ eprint = {arXiv:2212.12794},
eprinttype = {arxiv},
- primaryclass = {physics},
- publisher = {{arXiv}},
doi = {10.48550/arXiv.2212.12794},
url = {http://arxiv.org/abs/2212.12794},
urldate = {2022-12-31},
abstract = {We introduce a machine-learning (ML)-based weather simulator--called "GraphCast"--which outperforms the most accurate deterministic operational medium-range weather forecasting system in the world, as well as all previous ML baselines. GraphCast is an autoregressive model, based on graph neural networks and a novel high-resolution multi-scale mesh representation, which we trained on historical weather data from the European Centre for Medium-Range Weather Forecasts (ECMWF)'s ERA5 reanalysis archive. It can make 10-day forecasts, at 6-hour time intervals, of five surface variables and six atmospheric variables, each at 37 vertical pressure levels, on a 0.25-degree latitude-longitude grid, which corresponds to roughly 25 x 25 kilometer resolution at the equator. Our results show GraphCast is more accurate than ECMWF's deterministic operational forecasting system, HRES, on 90.0\% of the 2760 variable and lead time combinations we evaluated. GraphCast also outperforms the most accurate previous ML-based weather forecasting model on 99.2\% of the 252 targets it reported. GraphCast can generate a 10-day forecast (35 gigabytes of data) in under 60 seconds on Cloud TPU v4 hardware. Unlike traditional forecasting methods, ML-based forecasting scales well with data: by training on bigger, higher quality, and more recent data, the skill of the forecasts can improve. Together these results represent a key step forward in complementing and improving weather modeling with ML, open new opportunities for fast, accurate forecasting, and help realize the promise of ML-based simulation in the physical sciences.},
- archiveprefix = {arXiv},
+ pubstate = {preprint},
keywords = {/unread,DeepMind,GNN,PDE,SciML,simulation,weather forecasting},
- file = {/home/johannes/Nextcloud/Zotero/Lam et al_2022_GraphCast.pdf;/home/johannes/Zotero/storage/8UD54ESE/2212.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Lam et al_2022_GraphCast.pdf;/Users/wasmer/Zotero/storage/8UD54ESE/2212.html}
}
@unpublished{langerRepresentationsMoleculesMaterials2021,
@@ -4299,13 +4660,12 @@
date = {2021-02-09},
eprint = {2003.12081},
eprinttype = {arxiv},
- primaryclass = {cond-mat, physics:physics},
+ eprintclass = {cond-mat, physics:physics},
url = {http://arxiv.org/abs/2003.12081},
urldate = {2021-05-13},
abstract = {Computational study of molecules and materials from first principles is a cornerstone of physics, chemistry, and materials science, but limited by the cost of accurate and precise simulations. In settings involving many simulations, machine learning can reduce these costs, often by orders of magnitude, by interpolating between reference simulations. This requires representations that describe any molecule or material and support interpolation. We comprehensively review and discuss current representations and relations between them, using a unified mathematical framework based on many-body functions, group averaging, and tensor products. For selected state-of-the-art representations, we compare energy predictions for organic molecules, binary alloys, and Al-Ga-In sesquioxides in numerical experiments controlled for data distribution, regression method, and hyper-parameter optimization.},
- archiveprefix = {arXiv},
keywords = {_tablet,ACE,BoB,BS,CM,descriptors,GPR,KRR,library,materials,MBTR,ML,models,MTP,review,SOAP,with-code},
- file = {/home/johannes/Nextcloud/Zotero/Langer et al_2021_Representations of molecules and materials for interpolation of.pdf;/home/johannes/Zotero/storage/5BG77UWY/2003.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Langer et al_2021_Representations of molecules and materials for interpolation of.pdf;/Users/wasmer/Zotero/storage/5BG77UWY/2003.html}
}
@article{langerRepresentationsMoleculesMaterials2022,
@@ -4326,7 +4686,7 @@
issue = {1},
langid = {english},
keywords = {_tablet,ACE,benchmarking,BoB,BS,CM,descriptor comparison,descriptors,GPR,KRR,library,materials,MBTR,ML,models,MTP,review,SOAP,with-code},
- file = {/home/johannes/Nextcloud/Zotero/Langer et al_2022_Representations of molecules and materials for interpolation of.pdf;/home/johannes/Zotero/storage/9RVUDSSX/s41524-022-00721-x.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Langer et al_2022_Representations of molecules and materials for interpolation of.pdf;/Users/wasmer/Zotero/storage/9RVUDSSX/s41524-022-00721-x.html}
}
@article{larsenAtomicSimulationEnvironment2017,
@@ -4344,7 +4704,7 @@
urldate = {2021-10-17},
abstract = {The atomic simulation environment (ASE) is a software package written in the Python programming language with the aim of setting up, steering, and analyzing atomistic simulations. In ASE, tasks are fully scripted in Python. The powerful syntax of Python combined with the NumPy array library make it possible to perform very complex simulation tasks. For example, a sequence of calculations may be performed with the use of a simple ‘for-loop’ construction. Calculations of energy, forces, stresses and other quantities are performed through interfaces to many external electronic structure codes or force fields using a uniform interface. On top of this calculator interface, ASE provides modules for performing many standard simulation tasks such as structure optimization, molecular dynamics, handling of constraints and performing nudged elastic band calculations.},
langid = {english},
- file = {/home/johannes/Nextcloud/Zotero/Larsen et al_2017_The atomic simulation environment—a Python library for working with atoms.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Larsen et al_2017_The atomic simulation environment—a Python library for working with atoms.pdf}
}
@online{LearningLJPotential,
@@ -4352,7 +4712,7 @@
url = {https://teoroo-pinn.readthedocs.io/en/latest/notebooks/Learn_LJ_potential.html},
urldate = {2021-05-13},
keywords = {ANN,MD,ML,MLP,notebook},
- file = {/home/johannes/Zotero/storage/VDHVAB3I/Learn_LJ_potential.html}
+ file = {/Users/wasmer/Zotero/storage/VDHVAB3I/Learn_LJ_potential.html}
}
@article{lehtolaAssessmentInitialGuesses2019,
@@ -4372,7 +4732,26 @@
urldate = {2022-05-17},
abstract = {Electronic structure calculations, such as in the Hartree–Fock or Kohn–Sham density functional approach, require an initial guess for the molecular orbitals. The quality of the initial guess has a significant impact on the speed of convergence of the self-consistent field (SCF) procedure. Popular choices for the initial guess include the one-electron guess from the core Hamiltonian, the extended Hückel method, and the superposition of atomic densities (SAD). Here, we discuss alternative guesses obtained from the superposition of atomic potentials (SAP), which is easily implementable even in real-space calculations. We also discuss a variant of SAD which produces guess orbitals by purification of the density matrix that could also be used in real-space calculations, as well as a parameter-free variant of the extended Hückel method, which resembles the SAP method and is easy to implement on top of existing SAD infrastructure. The performance of the core Hamiltonian, the SAD, and the SAP guesses as well as the extended Hückel variant is assessed in nonrelativistic calculations on a data set of 259 molecules ranging from the first to the fourth periods by projecting the guess orbitals onto precomputed, converged SCF solutions in single- to triple-ζ basis sets. It is shown that the proposed SAP guess is the best guess on average. The extended Hückel guess offers a good alternative, with less scatter in accuracy.},
keywords = {initial guess,SCF},
- file = {/home/johannes/Nextcloud/Zotero/Lehtola_2019_Assessment of Initial Guesses for Self-Consistent Field Calculations.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Lehtola_2019_Assessment of Initial Guesses for Self-Consistent Field Calculations.pdf}
+}
+
+@article{lehtolaManyRecentDensity2022,
+ title = {Many Recent Density Functionals Are Numerically Ill-Behaved},
+ author = {Lehtola, Susi and Marques, Miguel A. L.},
+ date = {2022-11-07},
+ journaltitle = {The Journal of Chemical Physics},
+ shortjournal = {J. Chem. Phys.},
+ volume = {157},
+ number = {17},
+ pages = {174114},
+ publisher = {{American Institute of Physics}},
+ issn = {0021-9606},
+ doi = {10.1063/5.0121187},
+ url = {https://aip.scitation.org/doi/10.1063/5.0121187},
+ urldate = {2023-01-20},
+ abstract = {Most computational studies in chemistry and materials science are based on the use of density functional theory. Although the exact density functional is unknown, several density functional approximations (DFAs) offer a good balance of affordable computational cost and semi-quantitative accuracy for applications. The development of DFAs still continues on many fronts, and several new DFAs aiming for improved accuracy are published every year. However, the numerical behavior of these DFAs is an often-overlooked problem. In this work, we look at all 592 DFAs for three-dimensional systems available in Libxc 5.2.2 and examine the convergence of the density functional total energy based on tabulated atomic Hartree–Fock wave functions. We show that several recent DFAs, including the celebrated SCAN family of functionals, show impractically slow convergence with typically used numerical quadrature schemes, making these functionals unsuitable both for routine applications and high-precision studies, as thousands of radial quadrature points may be required to achieve sub-μEh accurate total energies for these functionals, while standard quadrature grids like the SG-3 grid only contain [Math Processing Error] O(100) radial quadrature points. These results are both a warning to users to always check the sufficiency of the quadrature grid when adopting novel functionals, as well as a guideline to the theory community to develop better-behaved density functionals.},
+ keywords = {/unread,DFA,DFT,numerical,SCF},
+ file = {/Users/wasmer/Nextcloud/Zotero/Lehtola_Marques_2022_Many recent density functionals are numerically ill-behaved.pdf}
}
@article{lehtolaRecentDevelopmentsLibxc2018,
@@ -4390,7 +4769,7 @@
abstract = {libxc is a library of exchange–correlation functionals for density-functional theory. We are concerned with semi-local functionals (or the semi-local part of hybrid functionals), namely local-density approximations, generalized-gradient approximations, and meta-generalized-gradient approximations. Currently we include around 400 functionals for the exchange, correlation, and the kinetic energy, spanning more than 50 years of research. Moreover, libxc is by now used by more than 20 codes, not only from the atomic, molecular, and solid-state physics, but also from the quantum chemistry communities.},
langid = {english},
keywords = {DFT,library,with-code},
- file = {/home/johannes/Nextcloud/Zotero/Lehtola et al_2018_Recent developments in libxc — A comprehensive library of functionals for.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Lehtola et al_2018_Recent developments in libxc — A comprehensive library of functionals for.pdf}
}
@article{leiDesignAnalysisMachine2019,
@@ -4408,7 +4787,7 @@
urldate = {2022-07-05},
abstract = {In this work we explore the potential of a data-driven approach to the design of exchange-correlation (xc) functionals. The approach, inspired by convolutional filters in computer vision and surrogate functions from optimization, utilizes convolutions of the electron density to form a feature space to represent local electronic environments and neural networks to map the features to the exchange-correlation energy density. These features are orbital free, and provide a systematic route to including information at various length scales. This work shows that convolutional descriptors are theoretically capable of an exact representation of the electron density, and proposes Maxwell-Cartesian spherical harmonic kernels as a class of rotationally invariant descriptors for the construction of machine learned functionals. The approach is demonstrated using data from the B3LYP functional on a number of small molecules containing C, H, O, and N along with a neural network regression model. The machine learned functionals are compared to standard physical approximations and the accuracy is assessed for the absolute energy of each molecular system as well as formation energies. The results indicate that it is possible to reproduce the exchange-correlation portion of B3LYP formation energies to within chemical accuracy using orbital-free descriptors with a spatial extent of 0.2 Ã…. The findings provide empirical insight into the spatial range of electron exchange, and suggest that the combination of convolutional descriptors and machine learning regression models is a promising framework for xc functional design, although challenges remain in obtaining training data and generating models consistent with pseudopotentials.},
keywords = {B3LYP,CNN,DFT,grid-based descriptors,LDA,MCSH,ML,ML-DFA,ML-DFT,ML-ESM,prediction from density,VWN},
- file = {/home/johannes/Nextcloud/Zotero/Lei_Medford_2019_Design and analysis of machine learning exchange-correlation functionals via.pdf;/home/johannes/Zotero/storage/RNGY77UQ/Lei and Medford - 2019 - Design and analysis of machine learning exchange-c.pdf;/home/johannes/Zotero/storage/D2P5RDDM/PhysRevMaterials.3.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Lei_Medford_2019_Design and analysis of machine learning exchange-correlation functionals via.pdf;/Users/wasmer/Zotero/storage/RNGY77UQ/Lei and Medford - 2019 - Design and analysis of machine learning exchange-c.pdf;/Users/wasmer/Zotero/storage/D2P5RDDM/PhysRevMaterials.3.html}
}
@article{lejaeghereErrorEstimatesSolidState2014,
@@ -4427,8 +4806,7 @@
urldate = {2021-10-15},
abstract = {Predictions of observable properties by density-functional theory calculations (DFT) are used increasingly often by experimental condensed-matter physicists and materials engineers as data. These predictions are used to analyze recent measurements, or to plan future experiments in a rational way. Increasingly more experimental scientists in these fields therefore face the natural question: what is the expected error for such a first-principles prediction? Information and experience about this question is implicitly available in the computational community, scattered over two decades of literature. The present review aims to summarize and quantify this implicit knowledge. This eventually leads to a practical protocol that allows any scientist—experimental or theoretical—to determine justifiable error estimates for many basic property predictions, without having to perform additional DFT calculations. A central role is played by a large and diverse test set of crystalline solids, containing all ground-state elemental crystals (except most lanthanides). For several properties of each crystal, the difference between DFT results and experimental values is assessed. We discuss trends in these deviations and review explanations suggested in the literature. A prerequisite for such an error analysis is that different implementations of the same first-principles formalism provide the same predictions. Therefore, the reproducibility of predictions across several mainstream methods and codes is discussed too. A quality factor Δ expresses the spread in predictions from two distinct DFT implementations by a single number. To compare the PAW method to the highly accurate APW+lo approach, a code assessment of VASP and GPAW (PAW) with respect to WIEN2k (APW+lo) yields Δ-values of 1.9 and 3.3 meV/atom, respectively. In both cases the PAW potentials recommended by the respective codes have been used. These differences are an order of magnitude smaller than the typical difference with experiment, and therefore predictions by APW+lo and PAW are for practical purposes identical.},
keywords = {benchmarking,code comparison,density-functional theory,error estimate},
- annotation = {\_eprint: https://doi.org/10.1080/10408436.2013.772503},
- file = {/home/johannes/Nextcloud/Zotero/Lejaeghere et al_2014_Error Estimates for Solid-State Density-Functional Theory Predictions.pdf;/home/johannes/Zotero/storage/92BC3LBZ/10408436.2013.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Lejaeghere et al_2014_Error Estimates for Solid-State Density-Functional Theory Predictions.pdf;/Users/wasmer/Zotero/storage/92BC3LBZ/10408436.2013.html}
}
@unpublished{lewisLearningElectronDensities2021,
@@ -4437,13 +4815,12 @@
date = {2021-06-09},
eprint = {2106.05364},
eprinttype = {arxiv},
- primaryclass = {cond-mat, physics:physics},
+ eprintclass = {cond-mat, physics:physics},
url = {http://arxiv.org/abs/2106.05364},
urldate = {2021-06-29},
abstract = {We introduce a local machine-learning method for predicting the electron densities of periodic systems. The framework is based on a numerical, atom-centred auxiliary basis, which enables an accurate expansion of the all-electron density in a form suitable for learning isolated and periodic systems alike. We show that using this formulation the electron densities of metals, semiconductors and molecular crystals can all be accurately predicted using a symmetry-adapted Gaussian process regression model, properly adjusted for the non-orthogonal nature of the basis. These predicted densities enable the efficient calculation of electronic properties which present errors on the order of tens of meV/atom when compared to ab initio density-functional calculations. We demonstrate the key power of this approach by using a model trained on ice unit cells containing only 4 water molecules to predict the electron densities of cells containing up to 512 molecules, and see no increase in the magnitude of the errors of derived electronic properties when increasing the system size. Indeed, we find that these extrapolated derived energies are more accurate than those predicted using a direct machine-learning model.},
- archiveprefix = {arXiv},
keywords = {DFT,GPR,library,ML,ML-DFT,ML-ESM,models,molecules,molecules & solids,prediction of electron density,prediction of ground-state properties,Resolution of the identity,SA-GPR,SALTED,solids,with-code},
- file = {/home/johannes/Nextcloud/Zotero/false;/home/johannes/Nextcloud/Zotero/Lewis et al_2021_Learning electron densities in the condensed-phase.pdf;/home/johannes/Zotero/storage/IC2NJGYT/2106.html}
+ file = {/home/johannes/Nextcloud/Zotero/false;/Users/wasmer/Nextcloud/Zotero/Lewis et al_2021_Learning electron densities in the condensed-phase.pdf;/Users/wasmer/Zotero/storage/IC2NJGYT/2106.html}
}
@article{lewisLearningElectronDensities2021a,
@@ -4462,7 +4839,7 @@
urldate = {2022-08-22},
abstract = {We introduce a local machine-learning method for predicting the electron densities of periodic systems. The framework is based on a numerical, atom-centered auxiliary basis, which enables an accurate expansion of the all-electron density in a form suitable for learning isolated and periodic systems alike. We show that, using this formulation, the electron densities of metals, semiconductors, and molecular crystals can all be accurately predicted using symmetry-adapted Gaussian process regression models, properly adjusted for the nonorthogonal nature of the basis. These predicted densities enable the efficient calculation of electronic properties, which present errors on the order of tens of meV/atom when compared to ab initio density-functional calculations. We demonstrate the key power of this approach by using a model trained on ice unit cells containing only 4 water molecules to predict the electron densities of cells containing up to 512 molecules and see no increase in the magnitude of the errors of derived electronic properties when increasing the system size. Indeed, we find that these extrapolated derived energies are more accurate than those predicted using a direct machine-learning model. Finally, on heterogeneous data sets SALTED can predict electron densities with errors below 4\%.},
keywords = {_tablet,DFT,GPR,library,ML,ML-DFT,ML-ESM,models,molecules,molecules & solids,prediction of electron density,prediction of ground-state properties,Resolution of the identity,SA-GPR,SALTED,solids,with-code},
- file = {/home/johannes/Nextcloud/Zotero/Lewis et al_2021_Learning Electron Densities in the Condensed Phase.pdf;/home/johannes/Zotero/storage/S9FT2FEZ/acs.jctc.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Lewis et al_2021_Learning Electron Densities in the Condensed Phase.pdf;/Users/wasmer/Zotero/storage/S9FT2FEZ/acs.jctc.html}
}
@unpublished{liDeepNeuralNetwork2021,
@@ -4471,13 +4848,12 @@
date = {2021-04-08},
eprint = {2104.03786},
eprinttype = {arxiv},
- primaryclass = {cond-mat, physics:physics, physics:quant-ph},
+ eprintclass = {cond-mat, physics:physics, physics:quant-ph},
url = {http://arxiv.org/abs/2104.03786},
urldate = {2022-01-02},
abstract = {The marriage of density functional theory (DFT) and deep learning methods has the potential to revolutionize modern research of material science. Here we study the crucial problem of representing DFT Hamiltonian for crystalline materials of arbitrary configurations via deep neural network. A general framework is proposed to deal with the infinite dimensionality and covariance transformation of DFT Hamiltonian matrix in virtue of locality and use message passing neural network together with graph representation for deep learning. Our example study on graphene-based systems demonstrates that high accuracy (\$\textbackslash sim\$meV) and good transferability can be obtained for DFT Hamiltonian, ensuring accurate predictions of materials properties without DFT. The Deep Hamiltonian method provides a solution to the accuracy-efficiency dilemma of DFT and opens new opportunities to explore large-scale materials and physics.},
- archiveprefix = {arXiv},
keywords = {Condensed Matter - Disordered Systems and Neural Networks,Condensed Matter - Materials Science,Condensed Matter - Mesoscale and Nanoscale Physics,Physics - Computational Physics,Quantum Physics},
- file = {/home/johannes/Nextcloud/Zotero/Li et al_2021_Deep Neural Network Representation of Density Functional Theory Hamiltonian.pdf;/home/johannes/Zotero/storage/B7RUP7VH/2104.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Li et al_2021_Deep Neural Network Representation of Density Functional Theory Hamiltonian.pdf;/Users/wasmer/Zotero/storage/B7RUP7VH/2104.html}
}
@article{liDLHubSimplifyingPublication2021,
@@ -4496,7 +4872,7 @@
abstract = {Machine Learning (ML) has become a critical tool enabling new methods of analysis and driving deeper understanding of phenomena across scientific disciplines. There is a growing need for “learning systems†to support various phases in the ML lifecycle. While others have focused on supporting model development, training, and inference, few have focused on the unique challenges inherent in science, such as the need to publish and share models and to serve them on a range of available computing resources. In this paper, we present the Data and Learning Hub for science (DLHub), a learning system designed to support these use cases. Specifically, DLHub enables publication of models, with descriptive metadata, persistent identifiers, and flexible access control. It packages arbitrary models into portable servable containers, and enables low-latency, distributed serving of these models on heterogeneous compute resources. We show that DLHub supports low-latency model inference comparable to other model serving systems including TensorFlow Serving, SageMaker, and Clipper, and improved performance, by up to 95\%, with batching and memoization enabled. We also show that DLHub can scale to concurrently serve models on 500 containers. Finally, we describe five case studies that highlight the use of DLHub for scientific applications.},
langid = {english},
keywords = {DLHub,Learning systems,Machine learning,Model serving},
- file = {/home/johannes/Zotero/storage/9B88LYEZ/S0743731520303464.html}
+ file = {/Users/wasmer/Zotero/storage/9B88LYEZ/S0743731520303464.html}
}
@article{liKohnShamEquationsRegularizer2021,
@@ -4515,7 +4891,7 @@
urldate = {2022-07-07},
abstract = {Including prior knowledge is important for effective machine learning models in physics and is usually achieved by explicitly adding loss terms or constraints on model architectures. Prior knowledge embedded in the physics computation itself rarely draws attention. We show that solving the Kohn-Sham equations when training neural networks for the exchange-correlation functional provides an implicit regularization that greatly improves generalization. Two separations suffice for learning the entire one-dimensional H2 dissociation curve within chemical accuracy, including the strongly correlated region. Our models also generalize to unseen types of molecules and overcome self-interaction error.},
keywords = {autodiff,CNN,DFT,JAX,JAX-DFT,Kohn-Sham regularizer,ML,ML-DFA,ML-DFT,ML-ESM,molecules,original publication,prediction from density,regularization,RNN},
- file = {/home/johannes/Nextcloud/Zotero/Li et al_2021_Kohn-Sham Equations as Regularizer.pdf;/home/johannes/Zotero/storage/CAFV9KV8/Li et al_2021_Kohn-Sham Equations as Regularizer.pdf;/home/johannes/Zotero/storage/QQA9HJV3/Li et al. - 2021 - Kohn-Sham Equations as Regularizer Building Prior.gif;/home/johannes/Zotero/storage/2MCFRSEU/PhysRevLett.126.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Li et al_2021_Kohn-Sham Equations as Regularizer.pdf;/Users/wasmer/Zotero/storage/CAFV9KV8/Li et al_2021_Kohn-Sham Equations as Regularizer.pdf;/Users/wasmer/Zotero/storage/QQA9HJV3/Li et al. - 2021 - Kohn-Sham Equations as Regularizer Building Prior.gif;/Users/wasmer/Zotero/storage/2MCFRSEU/PhysRevLett.126.html}
}
@article{liMetallizationSuperconductivityDense2014,
@@ -4534,7 +4910,7 @@
urldate = {2021-10-21},
abstract = {Hydrogen sulfide (H2S) is a prototype molecular system and a sister molecule of water (H2O). The phase diagram of solid H2S at high pressures remains largely unexplored arising from the challenges in dealing with the pressure-induced weakening of S–H bond and larger atomic core difference between H and S. Metallization is yet achieved for H2O, but it was observed for H2S above 96 GPa. However, the metallic structure of H2S remains elusive, greatly impeding the understanding of its metallicity and the potential superconductivity. We have performed an extensive structural study on solid H2S at pressure ranges of 10–200 GPa through an unbiased structure prediction method based on particle swarm optimization algorithm. Besides the findings of candidate structures for nonmetallic phases IV and V, we are able to establish stable metallic structures violating an earlier proposal of elemental decomposition into sulfur and hydrogen [R. Rousseau, M. Boero, M. Bernasconi, M. Parrinello, and K. Terakura, Phys. Rev. Lett. 85, 1254 (2000)]. Our study unravels a superconductive potential of metallic H2S with an estimated maximal transition temperature of ∼80 K at 160 GPa, higher than those predicted for most archetypal hydrogen-containing compounds (e.g., SiH4, GeH4, etc.).},
keywords = {applications of DFT,DFT,master-thesis,superconductor},
- file = {/home/johannes/Nextcloud/Zotero/Li et al_2014_The metallization and superconductivity of dense hydrogen sulfide.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Li et al_2014_The metallization and superconductivity of dense hydrogen sulfide.pdf}
}
@article{lindmaaTheoreticalPredictionProperties2017,
@@ -4548,25 +4924,23 @@
abstract = {The prediction of ground state properties of atomistic systems is of vital importance in technological advances as well as in the physical sciences. Fundamentally, these predictions are based on a ...},
langid = {english},
keywords = {Coulomb matrix,descriptors,DFT,Ewald sum matrix,kinetic energy density,KRR,ML,models,PCA,prediction of ground-state properties,prediction of solid formation energy},
- file = {/home/johannes/Nextcloud/Zotero/Lindmaa_2017_Theoretical prediction of properties of atomistic systems.pdf;/home/johannes/Zotero/storage/TVX96NQ7/record.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Lindmaa_2017_Theoretical prediction of properties of atomistic systems.pdf;/Users/wasmer/Zotero/storage/TVX96NQ7/record.html}
}
-@misc{liptonTroublingTrendsMachine2018,
+@online{liptonTroublingTrendsMachine2018,
title = {Troubling {{Trends}} in {{Machine Learning Scholarship}}},
author = {Lipton, Zachary C. and Steinhardt, Jacob},
date = {2018-07-26},
number = {arXiv:1807.03341},
- eprint = {1807.03341},
+ eprint = {arXiv:1807.03341},
eprinttype = {arxiv},
- primaryclass = {cs, stat},
- publisher = {{arXiv}},
doi = {10.48550/arXiv.1807.03341},
url = {http://arxiv.org/abs/1807.03341},
urldate = {2022-06-27},
abstract = {Collectively, machine learning (ML) researchers are engaged in the creation and dissemination of knowledge about data-driven algorithms. In a given paper, researchers might aspire to any subset of the following goals, among others: to theoretically characterize what is learnable, to obtain understanding through empirically rigorous experiments, or to build a working system that has high predictive accuracy. While determining which knowledge warrants inquiry may be subjective, once the topic is fixed, papers are most valuable to the community when they act in service of the reader, creating foundational knowledge and communicating as clearly as possible. Recent progress in machine learning comes despite frequent departures from these ideals. In this paper, we focus on the following four patterns that appear to us to be trending in ML scholarship: (i) failure to distinguish between explanation and speculation; (ii) failure to identify the sources of empirical gains, e.g., emphasizing unnecessary modifications to neural architectures when gains actually stem from hyper-parameter tuning; (iii) mathiness: the use of mathematics that obfuscates or impresses rather than clarifies, e.g., by confusing technical and non-technical concepts; and (iv) misuse of language, e.g., by choosing terms of art with colloquial connotations or by overloading established technical terms. While the causes behind these patterns are uncertain, possibilities include the rapid expansion of the community, the consequent thinness of the reviewer pool, and the often-misaligned incentives between scholarship and short-term measures of success (e.g., bibliometrics, attention, and entrepreneurial opportunity). While each pattern offers a corresponding remedy (don't do it), we also discuss some speculative suggestions for how the community might combat these trends.},
- archiveprefix = {arXiv},
+ pubstate = {preprint},
keywords = {criticism,ML,research ethics,state of a field},
- file = {/home/johannes/Nextcloud/Zotero/Lipton_Steinhardt_2018_Troubling Trends in Machine Learning Scholarship.pdf;/home/johannes/Zotero/storage/HK89ZR8C/1807.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Lipton_Steinhardt_2018_Troubling Trends in Machine Learning Scholarship.pdf;/Users/wasmer/Zotero/storage/HK89ZR8C/1807.html}
}
@article{liuDensityEstimationUsing2021,
@@ -4582,7 +4956,7 @@
url = {https://www.pnas.org/doi/10.1073/pnas.2101344118},
urldate = {2022-07-08},
keywords = {density,General ML},
- file = {/home/johannes/Nextcloud/Zotero/Liu et al_2021_Density estimation using deep generative neural networks.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Liu et al_2021_Density estimation using deep generative neural networks.pdf}
}
@article{liuImprovingPerformanceLongRangeCorrected2017,
@@ -4601,7 +4975,7 @@
urldate = {2022-07-05},
abstract = {A machine-learning-based exchange-correlation functional is proposed for general-purpose density functional theory calculations. It is built upon the long-range-corrected Becke–Lee–Yang–Parr (LC–BLYP) functional, along with an embedded neural network which determines the value of the range-separation parameter μ for every individual system. The structure and the weights of the neural network are optimized with a reference data set containing 368 highly accurate thermochemical and kinetic energies. The newly developed functional (LC–BLYP–NN) achieves a balanced performance for a variety of energetic properties investigated. It largely improves the accuracy of atomization energies and heats of formation on which the original LC–BLYP with a fixed μ performs rather poorly. Meanwhile, it yields a similar or slightly compromised accuracy for ionization potentials, electron affinities, and reaction barriers, for which the original LC–BLYP works reasonably well. This work clearly highlights the potential usefulness of machine-learning techniques for improving density functional calculations.},
keywords = {autoencoder,BLYP,compositional descriptors,DFT,ML,ML-DFA,ML-DFT,ML-ESM,molecules,NN,prediction of Exc,used small dataset},
- file = {/home/johannes/Nextcloud/Zotero/Liu et al_2017_Improving the Performance of Long-Range-Corrected Exchange-Correlation.pdf;/home/johannes/Zotero/storage/76EWRKPT/acs.jpca.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Liu et al_2017_Improving the Performance of Long-Range-Corrected Exchange-Correlation.pdf;/Users/wasmer/Zotero/storage/76EWRKPT/acs.jpca.html}
}
@article{liUnderstandingMachinelearnedDensity2016,
@@ -4619,8 +4993,7 @@
abstract = {Machine learning (ML) is an increasingly popular statistical tool for analyzing either measured or calculated data sets. Here, we explore its application to a well-defined physics problem, investigating issues of how the underlying physics is handled by ML, and how self-consistent solutions can be found by limiting the domain in which ML is applied. The particular problem is how to find accurate approximate density functionals for the kinetic energy (KE) of noninteracting electrons. Kernel ridge regression is used to approximate the KE of non-interacting fermions in a one dimensional box as a functional of their density. The properties of different kernels and methods of cross-validation are explored, reproducing the physics faithfully in some cases, but not others. We also address how self-consistency can be achieved with information on only a limited electronic density domain. Accurate constrained optimal densities are found via a modified Euler-Lagrange constrained minimization of the machine-learned total energy, despite the poor quality of its functional derivative. A projected gradient descent algorithm is derived using local principal component analysis. Additionally, a sparse grid representation of the density can be used without degrading the performance of the methods. The implications for machine-learned density functional approximations are discussed. © 2015 Wiley Periodicals, Inc.},
langid = {english},
keywords = {_tablet,DFA,DFT,KRR,ML,ML-DFA,ML-DFT,ML-ESM,tutorial},
- annotation = {\_eprint: https://onlinelibrary.wiley.com/doi/pdf/10.1002/qua.25040},
- file = {/home/johannes/Nextcloud/Zotero/Li et al_2016_Understanding machine-learned density functionals.pdf;/home/johannes/Zotero/storage/ZPNDJ7AU/qua.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Li et al_2016_Understanding machine-learned density functionals.pdf;/Users/wasmer/Zotero/storage/ZPNDJ7AU/qua.html}
}
@article{liuSingleNNModifiedBehler2020,
@@ -4640,7 +5013,7 @@
urldate = {2021-06-24},
abstract = {In this article, we introduce the SingleNN, which is a modified version of the Behler–Parrinello Neural Network (BPNN) where the neural networks for the prediction of atomic energy for different elements are combined into a single network with shared weights. Using a data set containing Cu, Ge, Li, Mo, Ni, and Si, we demonstrate that SingleNN could achieve an accuracy that is on a par with BPNN for energy and force predictions. Furthermore, we demonstrate that SingleNN could learn a common transformation for the fingerprints of atoms to a latent space in which the atomic energies of the atoms are nearly linear. Using the common transformation, we could fit the data with new elements by changing only weights in the output layer in the neural network. In this way, with a moderate compromise in accuracy, we can speed up the training process significantly and potentially reduce the amount of training data needed.},
keywords = {ACSF,ANN,BPNN,BPSF,descriptors,ML,MLP,models,SingleNN,surrogate model},
- file = {/home/johannes/Nextcloud/Zotero/Liu_Kitchin_2020_SingleNN.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Liu_Kitchin_2020_SingleNN.pdf}
}
@thesis{lohOvercomingDataScarcity2021,
@@ -4655,30 +5028,134 @@
langid = {english},
keywords = {contrastive learning,Deep learning,invariance,ML,NN,photonic crystals,rec-by-ruess,Schrödinger equation,small data,SSL,surrogate data,target: bandstructure,target: DOS,target: potential,TISE,transfer learning},
annotation = {Accepted: 2022-02-07T15:28:01Z},
- file = {/home/johannes/Nextcloud/Zotero/Loh_2021_Overcoming Data Scarcity in Deep Learning of Scientific Problems.pdf;/home/johannes/Zotero/storage/BLKX75W6/140165.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Loh_2021_Overcoming Data Scarcity in Deep Learning of Scientific Problems.pdf;/Users/wasmer/Zotero/storage/BLKX75W6/140165.html}
}
-@misc{lopanitsynaModelingHighentropyTransitionmetal2022,
+@article{longAcceleratingApproachDesigning2021,
+ title = {An Accelerating Approach of Designing Ferromagnetic Materials via Machine Learning Modeling of Magnetic Ground State and {{Curie}} Temperature},
+ author = {Long, Teng and Fortunato, Nuno M. and Zhang, Yixuan and Gutfleisch, Oliver and Zhang, Hongbin},
+ date = {2021-04-03},
+ journaltitle = {Materials Research Letters},
+ volume = {9},
+ number = {4},
+ pages = {169--174},
+ publisher = {{Taylor \& Francis}},
+ issn = {null},
+ doi = {10.1080/21663831.2020.1863876},
+ url = {https://doi.org/10.1080/21663831.2020.1863876},
+ urldate = {2023-01-25},
+ abstract = {Magnetic materials have a plethora of applications from information technologies to energy harvesting. However, their functionalities are often limited by the magnetic ordering temperature. In this work, we performed random forest on the magnetic ground state and the Curie temperature (TC) to classify ferromagnetic and antiferromagnetic compounds and to predict the TC of the ferromagnets. The resulting accuracy is about 87\% for classification and 91\% for regression. When the trained model is applied to magnetic intermetallic materials in Materials Project, the accuracy is comparable. Our work paves the way to accelerate the discovery of new magnetic compounds for technological applications.},
+ keywords = {/unread,autoencoder,compositional descriptors,magnetic materials,ML,pgi-1 seminar,prediction of Curie temperature,prediction of magnetic ground state,random forest},
+ file = {/Users/wasmer/Nextcloud/Zotero/Long et al_2021_An accelerating approach of designing ferromagnetic materials via machine.pdf}
+}
+
+@article{longInverseDesignCrystal2022,
+ title = {Inverse Design of Crystal Structures for Multicomponent Systems},
+ author = {Long, Teng and Zhang, Yixuan and Fortunato, Nuno M. and Shen, Chen and Dai, Mian and Zhang, Hongbin},
+ date = {2022-06-01},
+ journaltitle = {Acta Materialia},
+ shortjournal = {Acta Materialia},
+ volume = {231},
+ pages = {117898},
+ issn = {1359-6454},
+ doi = {10.1016/j.actamat.2022.117898},
+ url = {https://www.sciencedirect.com/science/article/pii/S135964542200283X},
+ urldate = {2023-01-25},
+ abstract = {We developed an inverse design framework, constrained crystal deep convolutional generative adversarial networks (CCDCGAN), enabling automated generation of stable multicomponent crystal structures. Their formation energy can be optimized in the latent space based on reversible crystal images with continuous representation. After training by 52,615 crystal structures from Materials Project, the CCDCGAN model is able to generate new crystal structure. It is demonstrated that when 9160 crystal structures are generated by CCDCGAN, 8310 of them have not been reported in the training set. Detailed analysis on 15 selected materials systems of the generated structures reveals that unreported crystal structures below the convex hull can be discovered in 6 of them. Moreover, the generation efficiency can be further improved by considering extra hypothetical structures in the training. This paves the way to perform inverse design of multicomponent materials with possible multi-objective optimization.},
+ langid = {english},
+ keywords = {/unread,CCDCGAN,crystal structure,crystal structure prediction,GAN,generative models,inverse design,materials project,pgi-1 seminar,with-code},
+ file = {/Users/wasmer/Nextcloud/Zotero/Long et al_2022_Inverse design of crystal structures for multicomponent systems.pdf;/Users/wasmer/Zotero/storage/6PDR73FE/S135964542200283X.html}
+}
+
+@online{lopanitsynaModelingHighentropyTransitionmetal2022,
title = {Modeling High-Entropy Transition-Metal Alloys with Alchemical Compression},
author = {Lopanitsyna, Nataliya and Fraux, Guillaume and Springer, Maximilian A. and De, Sandip and Ceriotti, Michele},
date = {2022-12-26},
number = {arXiv:2212.13254},
- eprint = {2212.13254},
+ eprint = {arXiv:2212.13254},
eprinttype = {arxiv},
- primaryclass = {cond-mat, physics:physics},
- publisher = {{arXiv}},
doi = {10.48550/arXiv.2212.13254},
url = {http://arxiv.org/abs/2212.13254},
urldate = {2022-12-29},
abstract = {Alloys composed of several elements in roughly equimolar composition, often referred to as high-entropy alloys, have long been of interest for their thermodynamics and peculiar mechanical properties, and more recently for their potential application in catalysis. They are a considerable challenge to traditional atomistic modeling, and also to data-driven potentials that for the most part have memory footprint, computational effort and data requirements which scale poorly with the number of elements included. We apply a recently proposed scheme to compress chemical information in a lower-dimensional space, which reduces dramatically the cost of the model with negligible loss of accuracy, to build a potential that can describe 25 d-block transition metals. The model shows semi-quantitative accuracy for prototypical alloys, and is remarkably stable when extrapolating to structures outside its training set. We use this framework to study element segregation in a computational experiment that simulates an equimolar alloy of all 25 elements, mimicking the seminal experiments by Cantor et al., and use our observations on the short-range order relations between the elements to define a data-driven set of Hume-Rothery rules that can serve as guidance for alloy design. We conclude with a study of three prototypical alloys, CoCrFeMnNi, CoCrFeMoNi and IrPdPtRhRu, determining their stability and the short-range order behavior of their constituents.},
- archiveprefix = {arXiv},
- keywords = {ACE,alchemical,chemical species scaling problem,descriptors,dimensionality reduction,high-entropy alloys,MTP,PyTorch,SOAP,transition metals,with-code},
- file = {/home/johannes/Nextcloud/Zotero/Lopanitsyna et al_2022_Modeling high-entropy transition-metal alloys with alchemical compression.pdf;/home/johannes/Zotero/storage/QNGQ9AQD/2212.html}
+ pubstate = {preprint},
+ keywords = {_tablet,ACE,alchemical,chemical species scaling problem,descriptors,dimensionality reduction,high-entropy alloys,MTP,PyTorch,SOAP,transition metals,with-code},
+ file = {/Users/wasmer/Nextcloud/Zotero/Lopanitsyna et al_2022_Modeling high-entropy transition-metal alloys with alchemical compression.pdf;/Users/wasmer/Nextcloud/Zotero/Lopanitsyna et al_2022_Modeling high-entropy transition-metal alloys with alchemical compression2.pdf;/Users/wasmer/Zotero/storage/QNGQ9AQD/2212.html}
+}
+
+@book{lounisTheoryMagneticTransition2007,
+ title = {Theory of {{Magnetic Transition Metal Nanoclusters}} on {{Surfaces}}},
+ author = {Lounis, Samir},
+ date = {2007},
+ series = {Schriften Des {{Forschungszentrums Jülich}}. {{Reihe Materie}} Und {{Material}} / {{Matter}} and {{Materials}}},
+ number = {41},
+ publisher = {{Forschungszentrum Jülich GmbH Zentralbibliothek, Verlag}},
+ location = {{Jülich}},
+ abstract = {The question how magnetism behaves when the dimension of materials is reduced to increasingly smaller sizes has attracted much research and led to the development of the field of magnetic nanostructures. This research has been fueled by the technological potential of these systems for the field of high-density magnetic storage media and has been accelerated by the many novel experimental methods and techniques developed exhibiting atomic resolution. This thesis is motivated by the quest for the understanding and the exploration of complex magnetism provided by atomic scale magnetic clusters deposited on surfaces or embedded in the bulk. The nature of magnetism in these systems can be very rich, in that the properties depend on the atomic species, the cluster size, shape and symmetry or choice of the substrate. Small variations of the cluster parameter may change the properties dramatically. Particularly rich and particularly challenging for experiment and theory is the behavior of clusters with competing magnetic interactions either between the cluster atoms or between the cluster and the substrate. In both cases magnetic frustration can lead to non-collinear magnetic structures for which the magnetic quantization axis changes from atom to atom. This thesis sheds light onto these systems from a theoretical perspective. Use is made of the density functional theory (DFT), the most successful material specific theory for describing electronic and derived properties from first-principles. Acting within this framework, we have developed and implemented the treatment of non-collinear magnetism into the Jülich version of the full-potential Korringa-Kohn-Rostoker Green Function (KKR-GF) method. The KKR-GF method provides several advantages compared to other first-principles methods. Based on solving the Dyson equation it allows an elegant treatment of non-periodic systems such as impurities and clusters in bulk or on surfaces. Electronic, magnetic properties and the observables provided by experimental techniques such as x-ray, scanning tunneling microscopy and spectroscopy can be accessed with the KKR-GF method. Firstly, the method was applied to 3\$\textbackslash textit\{d\}\$ transition-metal clusters on different ferromagnetic surfaces. Different types of magnetic clusters where selected. Clusters of Fe, Co, Ni atoms are ferromagnetic and thus magnetically collinear. In order to investigate magnetic frustration due to competing interactions within the ad-cluster we considered a (001) oriented surface of \$\textbackslash textit\{fcc\}\$ metals, a topology which usually does not lead to non-collinear magnetism. We tuned the strength of the magnetic coupling between the ad-clusters and the ferromagnetic surface by varying the substrate from the case of Ni(001) with a rather weak hybridization of the Ni \$\textbackslash textit\{d\}\$-states with the adatom \$\textbackslash textit\{d\}\$-states to the case of Fe\$\_\{3ML\}\$/Cu(001) with a much stronger hybridization due to the larger extend of the Fe wavefunctions. On Ni(001), the interaction between the Cr- as well as the Mn-dimer adatoms is of antiferromagnetic nature, which is in competition with the interaction with the substrate atoms. If we allow the magnetism to be non-collinear, the moments rotate such the Cr-(Mn) adatom moments are aligned antiparallel to each other and are basically perpendicular to the substrate moments. However, the weak AF(FM) interaction with the substrate causes a slight tilting towards the substrate, leading to an angle of 94.2◦(72.6◦) instead of 90◦. After performing total energy calculations we find that for Cr-dimer the ground state is collinear whereas the Mn-dimer prefers the non-collinear configuration as ground state. The Heisenberg model is shown [...]},
+ isbn = {978-3-89336-501-2},
+ langid = {english},
+ pagetotal = {189},
+ keywords = {Dissertation (Univ.),Hochschulschrift,juKKR,KKR,magnetism,PGI-1/IAS-1,thesis},
+ file = {/Users/wasmer/Nextcloud/Zotero/Lounis_2007_Theory of Magnetic Transition Metal Nanoclusters on Surfaces.pdf}
+}
+
+@article{lunghiComputationalDesignMagnetic2022,
+ title = {Computational Design of Magnetic Molecules and Their Environment Using Quantum Chemistry, Machine Learning and Multiscale Simulations},
+ author = {Lunghi, Alessandro and Sanvito, Stefano},
+ date = {2022-11},
+ journaltitle = {Nature Reviews Chemistry},
+ shortjournal = {Nat Rev Chem},
+ volume = {6},
+ number = {11},
+ pages = {761--781},
+ publisher = {{Nature Publishing Group}},
+ issn = {2397-3358},
+ doi = {10.1038/s41570-022-00424-3},
+ url = {https://www.nature.com/articles/s41570-022-00424-3},
+ urldate = {2023-02-24},
+ abstract = {Having served as a playground for fundamental studies on the physics of d and f electrons for almost a century, magnetic molecules are now becoming increasingly important for technological applications, such as magnetic resonance, data storage, spintronics and quantum information. All of these applications require the preservation and control of spins in time, an ability hampered by the interaction with the environment, namely with other spins, conduction electrons, molecular vibrations and electromagnetic fields. Thus, the design of a novel magnetic molecule with tailored properties is a formidable task, which does not only concern its electronic structures but also calls for a deep understanding of the interaction among all the degrees of freedom at play. This Review describes how state-of-the-art ab initio computational methods, combined with data-driven approaches to materials modelling, can be integrated into a fully multiscale strategy capable of defining design rules for magnetic molecules.},
+ issue = {11},
+ langid = {english},
+ keywords = {Computational chemistry,Electronic structure,Magnetic materials,Magnetic properties and materials,Spintronics},
+ file = {/Users/wasmer/Nextcloud/Zotero/Lunghi_Sanvito_2022_Computational design of magnetic molecules and their environment using quantum.pdf}
+}
+
+@online{lunghiSurfingMultipleConformationproperty2019,
+ title = {Surfing Multiple Conformation-Property Landscapes via Machine Learning: {{Designing}} Magnetic Anisotropy},
+ shorttitle = {Surfing Multiple Conformation-Property Landscapes via Machine Learning},
+ author = {Lunghi, Alessandro and Sanvito, Stefano},
+ date = {2019-11-06},
+ number = {arXiv:1911.02263},
+ eprint = {arXiv:1911.02263},
+ eprinttype = {arxiv},
+ doi = {10.48550/arXiv.1911.02263},
+ url = {http://arxiv.org/abs/1911.02263},
+ urldate = {2023-02-23},
+ abstract = {The advent of computational statistical disciplines, such as machine learning, is leading to a paradigm shift in the way we conceive the design of new compounds. Today computational science does not only provide a sound understanding of experiments, but also can directly design the best compound for specific applications. This approach, known as reverse engineering, requires the construction of models able to efficiently predict continuous structure-property maps. Here we show that reverse engineering can be used to tune the magnetic properties of a single-ion molecular magnet in an automated intelligent fashion. We design a machine learning model to predict both the energy and magnetic properties as function of the chemical structure. Then, a particle-swarm optimization algorithm is used to explore the conformational landscapes in the search for new molecular structures leading to an enhanced magnetic anisotropy. We find that a 5\% change in one of the coordination angles leads to a 50\% increase in the anisotropy. Our approach paves the way for a machine-learning-driven exploration of the chemical space of general classes of magnetic materials. Most importantly, it can be applied to any structure-property relation and offers an effective way to automatically generate new materials with target properties starting from the knowledge of previously synthesized ones.},
+ pubstate = {preprint},
+ keywords = {_tablet,/unread,Condensed Matter - Materials Science,Physics - Computational Physics},
+ file = {/Users/wasmer/Nextcloud/Zotero/Lunghi_Sanvito_2019_Surfing multiple conformation-property landscapes via machine learning.pdf;/Users/wasmer/Zotero/storage/FQSQYUBP/1911.html}
+}
+
+@online{lysogorskiyActiveLearningStrategies2022,
+ title = {Active Learning Strategies for Atomic Cluster Expansion Models},
+ author = {Lysogorskiy, Yury and Bochkarev, Anton and Mrovec, Matous and Drautz, Ralf},
+ date = {2022-12-16},
+ number = {arXiv:2212.08716},
+ eprint = {arXiv:2212.08716},
+ eprinttype = {arxiv},
+ doi = {10.48550/arXiv.2212.08716},
+ url = {http://arxiv.org/abs/2212.08716},
+ urldate = {2023-01-20},
+ abstract = {The atomic cluster expansion (ACE) was proposed recently as a new class of data-driven interatomic potentials with a formally complete basis set. Since the development of any interatomic potential requires a careful selection of training data and thorough validation, an automation of the construction of the training dataset as well as an indication of a model's uncertainty are highly desirable. In this work, we compare the performance of two approaches for uncertainty indication of ACE models based on the D-optimality criterion and ensemble learning. While both approaches show comparable predictions, the extrapolation grade based on the D-optimality (MaxVol algorithm) is more computationally efficient. In addition, the extrapolation grade indicator enables an active exploration of new structures, opening the way to the automated discovery of rare-event configurations. We demonstrate that active learning is also applicable to explore local atomic environments from large-scale MD simulations.},
+ pubstate = {preprint},
+ keywords = {/unread,ACE,active learning,Condensed Matter - Materials Science,D-optimality,descriptors,ensemble learning,MD,molecular dynamics,uncertainty quantification},
+ file = {/Users/wasmer/Nextcloud/Zotero/Lysogorskiy et al_2022_Active learning strategies for atomic cluster expansion models.pdf;/Users/wasmer/Zotero/storage/67ZIBP4V/2212.html}
}
@article{lysogorskiyPerformantImplementationAtomic2021,
title = {Performant Implementation of the Atomic Cluster Expansion ({{PACE}}) and Application to Copper and Silicon},
- author = {Lysogorskiy, Yury and van der Oord, Cas and Bochkarev, Anton and Menon, Sarath and Rinaldi, Matteo and Hammerschmidt, Thomas and Mrovec, Matous and Thompson, Aidan and Csányi, Gábor and Ortner, Christoph and Drautz, Ralf},
+ author = {Lysogorskiy, Yury and family=Oord, given=Cas, prefix=van der, useprefix=false and Bochkarev, Anton and Menon, Sarath and Rinaldi, Matteo and Hammerschmidt, Thomas and Mrovec, Matous and Thompson, Aidan and Csányi, Gábor and Ortner, Christoph and Drautz, Ralf},
date = {2021-06-28},
journaltitle = {npj Computational Materials},
shortjournal = {npj Comput Mater},
@@ -4694,7 +5171,7 @@
issue = {1},
langid = {english},
keywords = {_tablet,ACE,C++,descriptors,library},
- file = {/home/johannes/Nextcloud/Zotero/Lysogorskiy et al_2021_Performant implementation of the atomic cluster expansion (PACE) and.pdf;/home/johannes/Zotero/storage/QVQD97QT/s41524-021-00559-9.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Lysogorskiy et al_2021_Performant implementation of the atomic cluster expansion (PACE) and.pdf;/Users/wasmer/Zotero/storage/QVQD97QT/s41524-021-00559-9.html}
}
@article{magalhaesDensityFunctionalTheory2017,
@@ -4712,7 +5189,7 @@
urldate = {2021-10-21},
abstract = {The effect of light-absorbing atmospheric particles on climate change has been incorporated into climate models, but the absence of brown carbon (BrC) in these models has been leading to significant differences between model predictions and measured data on radiative forcing. Also, little is known regarding the relationship between optical properties and chemical compositions of BrC. Thus, we have characterized the absorption properties of catechol and known heterogeneous ozonolysis products, with a theoretical approach based on density functional theory (DFT). While catechol presents a weak absorption maximum in the ultraviolet C (UVC) region, other polyaromatic derivatives present an absorption up to 6 times higher, with biphenyl-2,2′,3,3′-tetraol, biphenyl-3,3′,4,4′,5,5′-hexaol, and terphenyl-2′,3,3′,3″,4,4″-hexaol presenting the strongest absorption. Moreover, these derivatives now absorb in the ultraviolet B (UVB) and ultraviolet A (UVA) regions, which are types of actinic radiation in the ultraviolet (UV) region not filtered by atmosphere (contrary to UVC), with terphenyl molecules presenting the highest absorption maximum. Furthermore, the absorption efficiency of these compounds is potentiated in the condensed phase, such as cloud droplets, rain, fog, and water films, as a result of a higher degree of electron delocalization. This study provides reliable information regarding the absorption properties of BrC generated by catechol, which is essential for the development of accurate models of climate forcing.},
keywords = {applications of DFT,atmospheric chemistry,DFT,master-thesis},
- file = {/home/johannes/Nextcloud/Zotero/Magalhães et al_2017_Density Functional Theory Calculation of the Absorption Properties of Brown.pdf;/home/johannes/Zotero/storage/ZRG83Z75/acsearthspacechem.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Magalhães et al_2017_Density Functional Theory Calculation of the Absorption Properties of Brown.pdf;/Users/wasmer/Zotero/storage/ZRG83Z75/acsearthspacechem.html}
}
@book{MagnetismElectronicStructure,
@@ -4721,7 +5198,7 @@
urldate = {2022-06-18},
langid = {english},
keywords = {condensed matter,defects,DFT,magnetism},
- file = {/home/johannes/Nextcloud/Zotero/Magnetism and the Electronic Structure of Crystals.pdf;/home/johannes/Zotero/storage/QVJRNHRA/978-3-642-84411-9.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Magnetism and the Electronic Structure of Crystals.pdf;/Users/wasmer/Zotero/storage/QVJRNHRA/978-3-642-84411-9.html}
}
@book{majlisQuantumTheoryMagnetism2007,
@@ -4736,7 +5213,7 @@
isbn = {978-981-256-792-5 978-981-277-974-8},
langid = {english},
keywords = {condensed matter,graduate,magnetism,textbook},
- file = {/home/johannes/Nextcloud/Zotero/Majlis_2007_The Quantum Theory of Magnetism.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Majlis_2007_The Quantum Theory of Magnetism.pdf}
}
@article{margrafPureNonlocalMachinelearned2021,
@@ -4757,7 +5234,7 @@
issue = {1},
langid = {english},
annotation = {Bandiera\_abtest: a Cc\_license\_type: cc\_by Cg\_type: Nature Research Journals Primary\_atype: Research Subject\_term: Computational chemistry;Density functional theory;Method development;Molecular dynamics Subject\_term\_id: computational-chemistry;density-functional-theory;method-development;molecular-dynamics},
- file = {/home/johannes/Nextcloud/Zotero/Margraf_Reuter_2021_Pure non-local machine-learned density functional theory for electron.pdf;/home/johannes/Zotero/storage/RCFG2NBC/s41467-020-20471-y.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Margraf_Reuter_2021_Pure non-local machine-learned density functional theory for electron.pdf;/Users/wasmer/Zotero/storage/RCFG2NBC/s41467-020-20471-y.html}
}
@online{MARVELDistinguishedLecture,
@@ -4765,7 +5242,7 @@
url = {https://nccr-marvel.ch/events/marvel-distinguished-lecture-GeorgKresse},
urldate = {2021-05-13},
keywords = {Bayesian regression,ML,ML-FF,MLP,models},
- file = {/home/johannes/Zotero/storage/IYCD348P/marvel-distinguished-lecture-GeorgKresse.html}
+ file = {/Users/wasmer/Zotero/storage/IYCD348P/marvel-distinguished-lecture-GeorgKresse.html}
}
@article{marzariElectronicstructureMethodsMaterials2021,
@@ -4785,8 +5262,8 @@
abstract = {The accuracy and efficiency of electronic-structure methods to understand, predict and design the properties of materials has driven a new paradigm in research. Simulations can greatly accelerate the identification, characterization and optimization of materials, with this acceleration driven by continuous progress in theory, algorithms and hardware, and by adaptation of concepts and tools from computer science. Nevertheless, the capability to identify and characterize materials relies on the predictive accuracy of the underlying physical descriptions, and on the ability to capture the complexity of realistic systems. We provide here an overview of electronic-structure methods, of their application to the prediction of materials properties, and of the different strategies employed towards the broader goals of materials design and discovery.},
issue = {6},
langid = {english},
- keywords = {Electronic structure},
- file = {/home/johannes/Nextcloud/Zotero/Marzari et al_2021_Electronic-structure methods for materials design.pdf;/home/johannes/Zotero/storage/AKF7QEMC/s41563-021-01013-3.html}
+ keywords = {database generation,DFT,Electronic structure,ESM,magnetism,ML,review},
+ file = {/Users/wasmer/Nextcloud/Zotero/Marzari et al_2021_Electronic-structure methods for materials design.pdf;/Users/wasmer/Zotero/storage/AKF7QEMC/s41563-021-01013-3.html}
}
@inproceedings{mavropoulosKorringaKohnRostokerKKRGreen2006,
@@ -4806,7 +5283,7 @@
isbn = {3-00-017350-1},
keywords = {_tablet,FZJ,KKR,PGI-1/IAS-1},
annotation = {Johannes Grotendorst, Stefan Blügel, Dominik Marx (Editors)},
- file = {/home/johannes/Nextcloud/Zotero/Mavropoulos_Papanikolaou_2006_The Korringa-Kohn-Rostoker (KKR) Green function method I.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Mavropoulos_Papanikolaou_2006_The Korringa-Kohn-Rostoker (KKR) Green function method I.pdf}
}
@article{mazinInverseOccamRazor2022,
@@ -4827,7 +5304,7 @@
issue = {4},
langid = {english},
keywords = {philosophy of science,physics,rec-by-ghosh,skeptics},
- file = {/home/johannes/Nextcloud/Zotero/Mazin_2022_Inverse Occam’s razor.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Mazin_2022_Inverse Occam’s razor.pdf}
}
@article{medvedevDensityFunctionalTheory2017,
@@ -4842,7 +5319,7 @@
doi = {10.1126/science.aah5975},
url = {https://www.science.org/doi/10.1126/science.aah5975},
urldate = {2021-11-17},
- file = {/home/johannes/Nextcloud/Zotero/Medvedev et al_2017_Density functional theory is straying from the path toward the exact functional.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Medvedev et al_2017_Density functional theory is straying from the path toward the exact functional.pdf}
}
@article{mehtaHighbiasLowvarianceIntroduction2019,
@@ -4861,7 +5338,7 @@
abstract = {Machine Learning (ML) is one of the most exciting and dynamic areas of modern research and application. The purpose of this review is to provide an introduction to the core concepts and tools of machine learning in a manner easily understood and intuitive to physicists. The review begins by covering fundamental concepts in ML and modern statistics such as the bias–variance tradeoff, overfitting, regularization, generalization, and gradient descent before moving on to more advanced topics in both supervised and unsupervised learning. Topics covered in the review include ensemble models, deep learning and neural networks, clustering and data visualization, energy-based models (including MaxEnt models and Restricted Boltzmann Machines), and variational methods. Throughout, we emphasize the many natural connections between ML and statistical physics. A notable aspect of the review is the use of Python Jupyter notebooks to introduce modern ML/statistical packages to readers using physics-inspired datasets (the Ising Model and Monte-Carlo simulations of supersymmetric decays of proton–proton collisions). We conclude with an extended outlook discussing possible uses of machine learning for furthering our understanding of the physical world as well as open problems in ML where physicists may be able to contribute.},
langid = {english},
keywords = {general,ML,notebooks,physics,rec-by-ruess,review,with-code},
- file = {/home/johannes/Nextcloud/Zotero/Mehta et al_2019_A high-bias, low-variance introduction to Machine Learning for physicists.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Mehta et al_2019_A high-bias, low-variance introduction to Machine Learning for physicists.pdf}
}
@article{meredigCanMachineLearning2018,
@@ -4881,7 +5358,7 @@
urldate = {2022-01-03},
abstract = {Traditional machine learning (ML) metrics overestimate model performance for materials discovery. We introduce (1) leave-one-cluster-out cross-validation (LOCO CV) and (2) a simple nearest-neighbor benchmark to show that model performance in discovery applications strongly depends on the problem, data sampling, and extrapolation. Our results suggest that ML-guided iterative experimentation may outperform standard high-throughput screening for discovering breakthrough materials like high-Tc superconductors with ML.},
langid = {english},
- file = {/home/johannes/Nextcloud/Zotero/Meredig et al_2018_Can machine learning identify the next high-temperature superconductor.pdf;/home/johannes/Zotero/storage/9WFQM4EG/c8me00012c.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Meredig et al_2018_Can machine learning identify the next high-temperature superconductor.pdf;/Users/wasmer/Zotero/storage/9WFQM4EG/c8me00012c.html}
}
@article{merkerMachineLearningMagnetism2022,
@@ -4899,8 +5376,8 @@
urldate = {2022-10-17},
abstract = {The determination of magnetic structure poses a long-standing challenge in condensed matter physics and materials science. Experimental techniques such as neutron diffraction are resource-limited and require complex structure refinement protocols, while computational approaches such as first-principles density functional theory (DFT) need additional semi-empirical correction, and reliable prediction is still largely limited to collinear magnetism. Here, we present a machine learning model that aims to classify the magnetic structure by inputting atomic coordinates containing transition metal and rare earth elements. By building a Euclidean equivariant neural network that preserves the crystallographic symmetry, the magnetic structure (ferromagnetic, antiferromagnetic, and non-magnetic) and magnetic propagation vector (zero or non-zero) can be predicted with an average accuracy of 77.8\% and 73.6\%. In particular, a 91\% accuracy is reached when predicting no magnetic ordering even if the structure contains magnetic element(s). Our work represents one step forward to solving the grand challenge of full magnetic structure determination.},
langid = {english},
- keywords = {classification,e3nn,equivariant,magnetic moment,magnetic order,magnetism,ML},
- file = {/home/johannes/Nextcloud/Zotero/Merker et al_2022_Machine learning magnetism classifiers from atomic coordinates.pdf;/home/johannes/Zotero/storage/7UQX89UL/S258900422201464X.html}
+ keywords = {_tablet,classification,e3nn,equivariant,magnetic moment,magnetic order,magnetism,ML},
+ file = {/Users/wasmer/Nextcloud/Zotero/Merker et al_2022_Machine learning magnetism classifiers from atomic coordinates.pdf;/Users/wasmer/Zotero/storage/7UQX89UL/S258900422201464X.html}
}
@article{merkysPosterioriMetadataAutomated2017,
@@ -4919,7 +5396,7 @@
urldate = {2021-09-18},
abstract = {In order to make results of computational scientific research findable, accessible, interoperable and re-usable, it is necessary to decorate them with standardised metadata. However, there are a number of technical and practical challenges that make this process difficult to achieve in practice. Here the implementation of a protocol is presented to tag crystal structures with their computed properties, without the need of human intervention to curate the data. This protocol leverages the capabilities of AiiDA, an open-source platform to manage and automate scientific computational workflows, and the TCOD, an open-access database storing computed materials properties using a well-defined and exhaustive ontology. Based on these, the complete procedure to deposit computed data in the TCOD database is automated. All relevant metadata are extracted from the full provenance information that AiiDA tracks and stores automatically while managing the calculations. Such a protocol also enables reproducibility of scientific data in the field of computational materials science. As a proof of concept, the AiiDA–TCOD interface is used to deposit 170 theoretical structures together with their computed properties and their full provenance graphs, consisting in over 4600 AiiDA nodes.},
keywords = {AiiDA,DFT,metadata},
- file = {/home/johannes/Nextcloud/Zotero/Merkys et al_2017_A posteriori metadata from automated provenance tracking.pdf;/home/johannes/Zotero/storage/9ZIMVPJ8/s13321-017-0242-y.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Merkys et al_2017_A posteriori metadata from automated provenance tracking.pdf;/Users/wasmer/Zotero/storage/9ZIMVPJ8/s13321-017-0242-y.html}
}
@inproceedings{missierW3CPROVFamily2013,
@@ -4936,7 +5413,7 @@
urldate = {2021-10-17},
abstract = {Provenance, a form of structured metadata designed to record the origin or source of information, can be instrumental in deciding whether information is to be trusted, how it can be integrated with other diverse information sources, and how to establish attribution of information to authors throughout its history. The PROV set of specifications, produced by the World Wide Web Consortium (W3C), is designed to promote the publication of provenance information on the Web, and offers a basis for interoperability across diverse provenance management systems. The PROV provenance model is deliberately generic and domain-agnostic, but extension mechanisms are available and can be exploited for modelling specific domains. This tutorial provides an account of these specifications. Starting from intuitive and informal examples that present idiomatic provenance patterns, it progressively introduces the relational model of provenance along with the constraints model for validation of provenance documents, and concludes with example applications that show the extension points in use.},
isbn = {978-1-4503-1597-5},
- file = {/home/johannes/Nextcloud/Zotero/Missier et al_2013_The W3C PROV family of specifications for modelling provenance metadata.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Missier et al_2013_The W3C PROV family of specifications for modelling provenance metadata.pdf}
}
@book{molnarGlobalSurrogateInterpretable,
@@ -4946,7 +5423,7 @@
urldate = {2021-05-13},
abstract = {Machine learning algorithms usually operate as black boxes and it is unclear how they derived a certain decision. This book is a guide for practitioners to make machine learning decisions interpretable.},
keywords = {ML,surrogate model},
- file = {/home/johannes/Zotero/storage/8VNJWN2F/global.html}
+ file = {/Users/wasmer/Zotero/storage/8VNJWN2F/global.html}
}
@article{morawietzDensityFunctionalTheoryBasedNeural2013,
@@ -4965,7 +5442,7 @@
urldate = {2021-05-18},
abstract = {The fundamental importance of water for many chemical processes has motivated the development of countless efficient but approximate water potentials for large-scale molecular dynamics simulations, from simple empirical force fields to very sophisticated flexible water models. Accurate and generally applicable water potentials should fulfill a number of requirements. They should have a quality close to quantum chemical methods, they should explicitly depend on all degrees of freedom including all relevant many-body interactions, and they should be able to describe molecular dissociation and recombination. In this work, we present a high-dimensional neural network (NN) potential for water clusters based on density-functional theory (DFT) calculations, which is constructed using clusters containing up to 10 monomers and is in principle able to meet all these requirements. We investigate the reliability of specific parametrizations employing two frequently used generalized gradient approximation (GGA) exchange-correlation functionals, PBE and RPBE, as reference methods. We find that the binding energy errors of the NN potentials with respect to DFT are significantly lower than the typical uncertainties of DFT calculations arising from the choice of the exchange-correlation functional. Further, we examine the role of van der Waals interactions, which are not properly described by GGA functionals. Specifically, we incorporate the D3 scheme suggested by Grimme (J. Chem. Phys. 2010, 132, 154104) in our potentials and demonstrate that it can be applied to GGA-based NN potentials in the same way as to DFT calculations without modification. Our results show that the description of small water clusters provided by the RPBE functional is significantly improved if van der Waals interactions are included, while in case of the PBE functional, which is well-known to yield stronger binding than RPBE, van der Waals corrections lead to overestimated binding energies.},
keywords = {chemistry,DFT,ML,MLP,models,NNP,vdW},
- file = {/home/johannes/Nextcloud/Zotero/Morawietz_Behler_2013_A Density-Functional Theory-Based Neural Network Potential for Water Clusters.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Morawietz_Behler_2013_A Density-Functional Theory-Based Neural Network Potential for Water Clusters.pdf}
}
@article{morenoDeepLearningHohenbergKohn2020,
@@ -4983,7 +5460,7 @@
urldate = {2021-12-14},
abstract = {A striking consequence of the Hohenberg-Kohn theorem of density functional theory is the existence of a bijection between the local density and the ground-state many-body wave function. Here we study the problem of constructing approximations to the Hohenberg-Kohn map using a statistical learning approach. Using supervised deep learning with synthetic data, we show that this map can be accurately constructed for a chain of one-dimensional interacting spinless fermions in different phases of this model including the charge ordered Mott insulator and metallic phases and the critical point separating them. However, we also find that the learning is less effective across quantum phase transitions, suggesting an intrinsic difficulty in efficiently learning nonsmooth functional relations. We further study the problem of directly reconstructing complex observables from simple local density measurements, proposing a scheme amenable to statistical learning from experimental data.},
keywords = {DFT,dunno,ML,ML-DFT,ML-ESM,Mott insulator,rec-by-bluegel},
- file = {/home/johannes/Nextcloud/Zotero/Moreno et al_2020_Deep Learning the Hohenberg-Kohn Maps of Density Functional Theory.pdf;/home/johannes/Zotero/storage/LWQ2IF97/Moreno et al_2020_Deep Learning the Hohenberg-Kohn Maps of Density Functional Theory2.pdf;/home/johannes/Zotero/storage/BCHNQKQ9/PhysRevLett.125.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Moreno et al_2020_Deep Learning the Hohenberg-Kohn Maps of Density Functional Theory.pdf;/Users/wasmer/Zotero/storage/LWQ2IF97/Moreno et al_2020_Deep Learning the Hohenberg-Kohn Maps of Density Functional Theory2.pdf;/Users/wasmer/Zotero/storage/BCHNQKQ9/PhysRevLett.125.html}
}
@article{morenoMachineLearningBand2021,
@@ -4996,7 +5473,7 @@
number = {8},
doi = {10.1103/PhysRevMaterials.5.083802},
keywords = {BPNN,DFT,ML,models,prediction from density,prediction of bandgap,rec-by-kim},
- file = {/home/johannes/Nextcloud/Zotero/Moreno_2021_Machine learning band gaps from the electron density.pdf;/home/johannes/Zotero/storage/B9EJXFVY/PhysRevMaterials.5.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Moreno_2021_Machine learning band gaps from the electron density.pdf;/Users/wasmer/Zotero/storage/B9EJXFVY/PhysRevMaterials.5.html}
}
@article{morganOpportunitiesChallengesMachine2020,
@@ -5014,7 +5491,7 @@
urldate = {2021-06-29},
abstract = {Advances in machine learning have impacted myriad areas of materials science, such as the discovery of novel materials and the improvement of molecular simulations, with likely many more important developments to come. Given the rapid changes in this field, it is challenging to understand both the breadth of opportunities and the best practices for their use. In this review, we address aspects of both problems by providing an overview of the areas in which machine learning has recently had significant impact in materials science, and then we provide a more detailed discussion on determining the accuracy and domain of applicability of some common types of machine learning models. Finally, we discuss some opportunities and challenges for the materials community to fully utilize the capabilities of machine learning.},
langid = {english},
- file = {/home/johannes/Nextcloud/Zotero/Morgan_Jacobs_2020_Opportunities and Challenges for Machine Learning in Materials Science.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Morgan_Jacobs_2020_Opportunities and Challenges for Machine Learning in Materials Science.pdf}
}
@article{morgensternStrongWeak3D2021,
@@ -5032,26 +5509,23 @@
abstract = {The contributions of surface science methods to discover and improve 3D topological insulator materials are reviewed herein, illustrated with examples from the authors’ own work. In particular, it is demonstrated that spin-polarized angular-resolved photoelectron spectroscopy is instrumental to evidence the spin-helical surface Dirac cone, to tune its Dirac point energy toward the Fermi level, and to discover novel types of topological insulators such as dual ones or switchable ones in phase change materials. Moreover, procedures are introduced to spatially map potential fluctuations by scanning tunneling spectroscopy and to identify topological edge states in weak topological insulators.},
langid = {english},
keywords = {angular-resolved photoelectron spectroscopy,scanning tunneling spectroscopy,spin-polarized topological insulators},
- annotation = {\_eprint: https://onlinelibrary.wiley.com/doi/pdf/10.1002/pssb.202000060},
- file = {/home/johannes/Nextcloud/Zotero/Morgenstern et al_2021_Strong and Weak 3D Topological Insulators Probed by Surface Science Methods.pdf;/home/johannes/Zotero/storage/4RCNJ2RK/pssb.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Morgenstern et al_2021_Strong and Weak 3D Topological Insulators Probed by Surface Science Methods.pdf;/Users/wasmer/Zotero/storage/4RCNJ2RK/pssb.html}
}
-@misc{morrowHowValidateMachinelearned2022,
+@online{morrowHowValidateMachinelearned2022,
title = {How to Validate Machine-Learned Interatomic Potentials},
author = {Morrow, Joe D. and Gardner, John L. A. and Deringer, Volker L.},
date = {2022-11-28},
number = {arXiv:2211.12484},
- eprint = {2211.12484},
+ eprint = {arXiv:2211.12484},
eprinttype = {arxiv},
- primaryclass = {cond-mat, physics:physics},
- publisher = {{arXiv}},
doi = {10.48550/arXiv.2211.12484},
url = {http://arxiv.org/abs/2211.12484},
urldate = {2023-01-02},
abstract = {Machine learning (ML) approaches enable large-scale atomistic simulations with near-quantum-mechanical accuracy. With the growing availability of these methods there arises a need for careful validation, particularly for physically agnostic models - that is, for potentials which extract the nature of atomic interactions from reference data. Here, we review the basic principles behind ML potentials and their validation for atomic-scale materials modeling. We discuss best practice in defining error metrics based on numerical performance as well as physically guided validation. We give specific recommendations that we hope will be useful for the wider community, including those researchers who intend to use ML potentials for materials "off the shelf".},
- archiveprefix = {arXiv},
- keywords = {benchmarking,best practices,how-to,ML,MLP,tutorial},
- file = {/home/johannes/Nextcloud/Zotero/Morrow et al_2022_How to validate machine-learned interatomic potentials.pdf;/home/johannes/Zotero/storage/TW3TCHB3/2211.html}
+ pubstate = {preprint},
+ keywords = {_tablet,benchmarking,best practices,cross-validation,experimental data,GAP,how-to,ML,MLP,model evaluation,model validation,models comparison,MTP,numerical errors,SOAP,SOTA,tutorial},
+ file = {/Users/wasmer/Nextcloud/Zotero/Morrow et al_2022_How to validate machine-learned interatomic potentials.pdf;/Users/wasmer/Zotero/storage/TW3TCHB3/2211.html}
}
@thesis{mozumderDesignMagneticInteractions2022,
@@ -5063,8 +5537,8 @@
abstract = {Magnetic impurities and their long-range interaction (ferromagnetic) order play pivotal roles in the topological phase transition from QSHI to QAHI. This transition transforms helical edge states belonging to the QSHI for 2D TIs (surface states for 3D TI) to the chiral edge states in QAHI for 2D TIs (surface states for 3D TI). Due to such chiral states, the QAHIs forbid back-scattering in electron conducting channels, which in turn provide passionless current and increase energy efficiency for conduct- ing channels. The chiral states are consist of single spin electrons which provide spin currents from conventional charge currents. Regarding the properties of QAHIs, the QAHIs opens a new venue for low-energy elec- tronics, spintronics and quantum computation [9]. Independently, the V- [10] and Cr-doped [11] as well as their co-doping [12] (Sb, Bi)2 Te3 shows stable QAHE but with very low temperatures (≤ 0.5K). In this high throughput ab-initio work, we will investigate other possible co-doping, dimer calculations, from the d-block elements in 3D TI Bi2 Te3 . For this purpose, we have extended AiiDA-KKR plugins by developing combine- impurity workflow called combine imps wc using GF formulation of DFT code (KKR-GF method) and the new workflow is capable to run multi- impurity calculations. Here, the dimer calculations are in the main fo- cus, and from the calculation results we will analyze Heisenberg isotropic collinear interaction (Jij ), Dzyaloshinskii–Moriya interaction (DMI, Dij ), and their ratio for each possible impurity couple. Finally, using the ob- tained Jij data we have implemented some linear regression machine learn- ing tools to understand better the dependency of Jij on some well-known factors e.g. inter-impurity distance, electronegativity. Our results from the notion of this work will give a list of some potential impurities and after their potential impurity combinations for stable QAHE. It will also render an impression of implementation of machine learning approach for designing better magnetic interactions in TIs.},
langid = {english},
pagetotal = {85},
- keywords = {_tablet,master-thesis,PGI-1/IAS-1,thesis},
- file = {/home/johannes/Nextcloud/Zotero/Mozumder_2022_Design of magnetic interactions in doped topological insulators.pdf}
+ keywords = {_tablet,AiiDA,aiida-kkr,Hall QAHE,impurity embedding,juKKR,KKR,master-thesis,PGI-1/IAS-1,thesis},
+ file = {/Users/wasmer/Nextcloud/Zotero/Mozumder_2022_Design of magnetic interactions in doped topological insulators.pdf}
}
@article{mullerSpiritMultifunctionalFramework2019,
@@ -5078,7 +5552,7 @@
number = {22},
doi = {10.1103/PhysRevB.99.224414},
keywords = {browser-based visualization,interactive visualization,library,PGI-1/IAS-1,spin dynamics,Spirit,visualization,web app,with-code},
- file = {/home/johannes/Nextcloud/Zotero/Müller_2019_iSpirit-i.pdf;/home/johannes/Zotero/storage/NXE55BTB/PhysRevB.99.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Müller_2019_iSpirit-i.pdf;/Users/wasmer/Zotero/storage/NXE55BTB/PhysRevB.99.html}
}
@book{MultipleScatteringTheory,
@@ -5088,7 +5562,7 @@
urldate = {2021-12-02},
isbn = {978-0-7503-1490-9},
langid = {english},
- file = {/home/johannes/Nextcloud/Zotero/Multiple Scattering Theory.pdf;/home/johannes/Zotero/storage/UYLUXULV/978-0-7503-1490-9.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Multiple Scattering Theory.pdf;/Users/wasmer/Zotero/storage/UYLUXULV/978-0-7503-1490-9.html}
}
@unpublished{musaelianLearningLocalEquivariant2022,
@@ -5097,13 +5571,12 @@
date = {2022-04-11},
eprint = {2204.05249},
eprinttype = {arxiv},
- primaryclass = {cond-mat, physics:physics},
+ eprintclass = {cond-mat, physics:physics},
url = {http://arxiv.org/abs/2204.05249},
urldate = {2022-04-14},
abstract = {A simultaneously accurate and computationally efficient parametrization of the energy and atomic forces of molecules and materials is a long-standing goal in the natural sciences. In pursuit of this goal, neural message passing has lead to a paradigm shift by describing many-body correlations of atoms through iteratively passing messages along an atomistic graph. This propagation of information, however, makes parallel computation difficult and limits the length scales that can be studied. Strictly local descriptor-based methods, on the other hand, can scale to large systems but do not currently match the high accuracy observed with message passing approaches. This work introduces Allegro, a strictly local equivariant deep learning interatomic potential that simultaneously exhibits excellent accuracy and scalability of parallel computation. Allegro learns many-body functions of atomic coordinates using a series of tensor products of learned equivariant representations, but without relying on message passing. Allegro obtains improvements over state-of-the-art methods on the QM9 and revised MD-17 data sets. A single tensor product layer is shown to outperform existing deep message passing neural networks and transformers on the QM9 benchmark. Furthermore, Allegro displays remarkable generalization to out-of-distribution data. Molecular dynamics simulations based on Allegro recover structural and kinetic properties of an amorphous phosphate electrolyte in excellent agreement with first principles calculations. Finally, we demonstrate the parallel scaling of Allegro with a dynamics simulation of 100 million atoms.},
- archiveprefix = {arXiv},
keywords = {Allegro,GNN,MD,ML,MLP,MPNN,NequIP},
- file = {/home/johannes/Nextcloud/Zotero/Musaelian et al_2022_Learning Local Equivariant Representations for Large-Scale Atomistic Dynamics.pdf;/home/johannes/Zotero/storage/3GTGKKHF/2204.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Musaelian et al_2022_Learning Local Equivariant Representations for Large-Scale Atomistic Dynamics.pdf;/Users/wasmer/Zotero/storage/3GTGKKHF/2204.html}
}
@article{musilEfficientImplementationAtomdensity2021,
@@ -5122,7 +5595,7 @@
urldate = {2021-05-13},
abstract = {Physically motivated and mathematically robust atom-centered representations of molecular structures are key to the success of modern atomistic machine learning. They lie at the foundation of a wide range of methods to predict the properties of both materials and molecules and to explore and visualize their chemical structures and compositions. Recently, it has become clear that many of the most effective representations share a fundamental formal connection. They can all be expressed as a discretization of n-body correlation functions of the local atom density, suggesting the opportunity of standardizing and, more importantly, optimizing their evaluation. We present an implementation, named librascal, whose modular design lends itself both to developing refinements to the density-based formalism and to rapid prototyping for new developments of rotationally equivariant atomistic representations. As an example, we discuss smooth overlap of atomic position (SOAP) features, perhaps the most widely used member of this family of representations, to show how the expansion of the local density can be optimized for any choice of radial basis sets. We discuss the representation in the context of a kernel ridge regression model, commonly used with SOAP features, and analyze how the computational effort scales for each of the individual steps of the calculation. By applying data reduction techniques in feature space, we show how to reduce the total computational cost by a factor of up to 4 without affecting the model’s symmetry properties and without significantly impacting its accuracy.},
keywords = {descriptors,library,librascal,ML,SOAP,with-code},
- file = {/home/johannes/Nextcloud/Zotero/Musil et al_2021_Efficient implementation of atom-density representations.pdf;/home/johannes/Zotero/storage/A3DVYDNI/5.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Musil et al_2021_Efficient implementation of atom-density representations.pdf;/Users/wasmer/Zotero/storage/A3DVYDNI/5.html}
}
@article{musilFastAccurateUncertainty2019,
@@ -5141,7 +5614,7 @@
urldate = {2021-05-30},
abstract = {We present a scheme to obtain an inexpensive and reliable estimate of the uncertainty associated with the predictions of a machine-learning model of atomic and molecular properties. The scheme is based on resampling, with multiple models being generated based on subsampling of the same training data. The accuracy of the uncertainty prediction can be benchmarked by maximum likelihood estimation, which can also be used to correct for correlations between resampled models and to improve the performance of the uncertainty estimation by a cross-validation procedure. In the case of sparse Gaussian Process Regression models, this resampled estimator can be evaluated at negligible cost. We demonstrate the reliability of these estimates for the prediction of molecular and materials energetics and for the estimation of nuclear chemical shieldings in molecular crystals. Extension to estimate the uncertainty in energy differences, forces, or other correlated predictions is straightforward. This method can be easily applied to other machine-learning schemes and will be beneficial to make data-driven predictions more reliable and to facilitate training-set optimization and active-learning strategies.},
keywords = {descriptors,GPR,library,ML,models,SA-GPR,SOAP,uncertainty quantification,with-code},
- file = {/home/johannes/Nextcloud/Zotero/Musil et al_2019_Fast and Accurate Uncertainty Estimation in Chemical Machine Learning.pdf;/home/johannes/Zotero/storage/PGUZKGX5/acs.jctc.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Musil et al_2019_Fast and Accurate Uncertainty Estimation in Chemical Machine Learning.pdf;/Users/wasmer/Zotero/storage/PGUZKGX5/acs.jctc.html}
}
@article{musilMachineLearningAtomic2019,
@@ -5160,7 +5633,7 @@
abstract = {Statistical learning algorithms are finding more and more applications in science and technology. Atomic-scale modeling is no exception, with machine learning becoming commonplace as a tool to predict energy, forces and properties of molecules and condensed-phase systems. This short review summarizes recent progress in the field, focusing in particular on the problem of representing an atomic configuration in a mathematically robust and computationally efficient way. We also discuss some of the regression algorithms that have been used to construct surrogate models of atomic-scale properties. We then show examples of how the optimization of the machine-learning models can both incorporate and reveal insights onto the physical phenomena that underlie structure–property relations.},
langid = {english},
keywords = {descriptors,descriptors analysis,ML,SOAP},
- file = {/home/johannes/Nextcloud/Zotero/Musil_Ceriotti_2019_Machine Learning at the Atomic Scale.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Musil_Ceriotti_2019_Machine Learning at the Atomic Scale.pdf}
}
@article{musilPhysicsInspiredStructuralRepresentations2021,
@@ -5176,7 +5649,7 @@
urldate = {2021-08-16},
abstract = {The first step in the construction of a regression model or a data-driven analysis, aiming to predict or elucidate the relationship between the atomic-scale structure of matter and its properties, involves transforming the Cartesian coordinates of the atoms into a suitable representation. The development of atomic-scale representations has played, and continues to play, a central role in the success of machine-learning methods for chemistry and materials science. This review summarizes the current understanding of the nature and characteristics of the most commonly used structural and chemical descriptions of atomistic structures, highlighting the deep underlying connections between different frameworks and the ideas that lead to computationally efficient and universally applicable models. It emphasizes the link between properties, structures, their physical chemistry, and their mathematical description, provides examples of recent applications to a diverse set of chemical and materials science problems, and outlines the open questions and the most promising research directions in the field.},
keywords = {_tablet,ACSF,descriptor comparison,descriptor dimred,descriptors,MBTR,ML,review,SOAP},
- file = {/home/johannes/Nextcloud/Zotero/Musil et al_2021_Physics-Inspired Structural Representations for Molecules and Materials.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Musil et al_2021_Physics-Inspired Structural Representations for Molecules and Materials.pdf}
}
@unpublished{musilPhysicsinspiredStructuralRepresentations2021,
@@ -5185,13 +5658,12 @@
date = {2021-05-04},
eprint = {2101.04673},
eprinttype = {arxiv},
- primaryclass = {physics},
+ eprintclass = {physics},
url = {http://arxiv.org/abs/2101.04673},
urldate = {2021-05-30},
abstract = {The first step in the construction of a regression model or a data-driven analysis framework for matter at the atomic scale involves transforming the Cartesian coordinates that describe the positions of the atoms in the form of a representation that obeys the same symmetries as the properties of interest, and in general reflects the physical nature of the problem. The link between properties, structures, their physical chemistry and their mathematical description is strongest when it comes to applications aimed at determining a precise correspondence between atomic configurations and the quantities that one might compute by a quantum mechanical electronic-structure calculation or measure experimentally. The development of atomic-scale representations have played, and continue to play, a central role in the success of machine-learning methods that rely on this correspondence, such as interatomic potentials, as well as generic property models, structural classifiers and low-dimensional maps of structures and datasets. This review summarizes the current understanding of the nature and characteristics of the most commonly used structural and chemical descriptions of molecules and materials, highlighting the deep underlying connections between different frameworks, and the ideas that lead to computationally efficient and universally applicable models. It gives examples of recent applications to a diverse set of chemical and materials science problems, and outlines the open questions and the most promising research directions in the field.},
- archiveprefix = {arXiv},
keywords = {_tablet,ACSF,descriptor dimred,descriptors,descriptors analysis,MBTR,ML,review,SOAP},
- file = {/home/johannes/Nextcloud/Zotero/Musil et al_2021_Physics-inspired structural representations for molecules and materials.pdf;/home/johannes/Zotero/storage/EXTUHGNH/2101.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Musil et al_2021_Physics-inspired structural representations for molecules and materials.pdf;/Users/wasmer/Zotero/storage/EXTUHGNH/2101.html}
}
@article{nagaosaTopologicalPropertiesDynamics2013,
@@ -5212,7 +5684,7 @@
issue = {12},
langid = {english},
annotation = {Bandiera\_abtest: a Cg\_type: Nature Research Journals Primary\_atype: Reviews Subject\_term: Magnetic properties and materials Subject\_term\_id: magnetic-properties-and-materials},
- file = {/home/johannes/Nextcloud/Zotero/Nagaosa_Tokura_2013_Topological properties and dynamics of magnetic skyrmions.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Nagaosa_Tokura_2013_Topological properties and dynamics of magnetic skyrmions.pdf}
}
@article{narayanAssessingSinglecellTranscriptomic2021,
@@ -5230,7 +5702,7 @@
abstract = {Nonlinear data visualization methods, such as t-distributed stochastic neighbor embedding (t-SNE) and uniform manifold approximation and projection (UMAP), summarize the complex transcriptomic landscape of single cells in two dimensions or three dimensions, but they neglect the local density of data points in the original space, often resulting in misleading visualizations where densely populated subsets of cells are given more visual space than warranted by their transcriptional diversity in the dataset. Here we present den-SNE and densMAP, which are density-preserving visualization tools based on t-SNE and UMAP, respectively, and demonstrate their ability to accurately incorporate information about transcriptomic variability into the visual interpretation of single-cell RNA sequencing data. Applied to recently published datasets, our methods reveal significant changes in transcriptomic variability in a range of biological processes, including heterogeneity in transcriptomic variability of immune cells in blood and tumor, human immune cell specialization and the developmental trajectory of Caenorhabditis elegans. Our methods are readily applicable to visualizing high-dimensional data in other scientific domains.},
langid = {english},
keywords = {den-SNE,density-preserving,densMAP,dimensionality reduction,library,t-SNE,UMAP,unsupervised learning,visualization,with-code},
- file = {/home/johannes/Nextcloud/Zotero/Narayan et al_2021_Assessing single-cell transcriptomic variability through density-preserving.pdf;/home/johannes/Zotero/storage/72VGP9LG/s41587-020-00801-7.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Narayan et al_2021_Assessing single-cell transcriptomic variability through density-preserving.pdf;/Users/wasmer/Zotero/storage/72VGP9LG/s41587-020-00801-7.html}
}
@article{narayanDensityPreservingDataVisualization2020,
@@ -5246,7 +5718,88 @@
abstract = {{$<$}p{$>$}Nonlinear data-visualization methods, such as t-SNE and UMAP, have become staple tools for summarizing the complex transcriptomic landscape of single cells in 2D or 3D. However, existing approaches neglect the local density of data points in the original space, often resulting in misleading visualizations where densely populated subpopulations of cells are given more visual space even if they account for only a small fraction of transcriptional diversity within the dataset. We present den-SNE and densMAP, our density-preserving visualization tools based on t-SNE and UMAP, respectively, and demonstrate their ability to facilitate more accurate visual interpretation of single-cell RNA-seq data. On recently published datasets, our methods newly reveal significant changes in transcriptomic variability within a range of biological processes, including cancer, immune cell specialization in human, and the developmental trajectory of \emph{C. elegans}. Our methods are readily applicable to visualizing high-dimensional data in other scientific domains.{$<$}/p{$>$}},
langid = {english},
keywords = {den-SNE,density-preserving,densMAP,dimensionality reduction,library,t-SNE,UMAP,unsupervised learning,visualization,with-code},
- file = {/home/johannes/Nextcloud/Zotero/Narayan et al_2020_Density-Preserving Data Visualization Unveils Dynamic Patterns of Single-Cell.pdf;/home/johannes/Zotero/storage/6QBY65KW/2020.05.12.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Narayan et al_2020_Density-Preserving Data Visualization Unveils Dynamic Patterns of Single-Cell.pdf;/Users/wasmer/Zotero/storage/6QBY65KW/2020.05.12.html}
+}
+
+@article{nelsonDataDrivenTimePropagation2022,
+ title = {Data-{{Driven Time Propagation}} of {{Quantum Systems}} with {{Neural Networks}}},
+ author = {Nelson, James and Coopmans, Luuk and Kells, Graham and Sanvito, Stefano},
+ date = {2022-07-01},
+ journaltitle = {Physical Review B},
+ shortjournal = {Phys. Rev. B},
+ volume = {106},
+ number = {4},
+ eprint = {2201.11647},
+ eprinttype = {arxiv},
+ eprintclass = {cond-mat, physics:quant-ph},
+ pages = {045402},
+ issn = {2469-9950, 2469-9969},
+ doi = {10.1103/PhysRevB.106.045402},
+ url = {http://arxiv.org/abs/2201.11647},
+ urldate = {2023-02-23},
+ abstract = {We investigate the potential of supervised machine learning to propagate a quantum system in time. While Markovian dynamics can be learned easily, given a sufficient amount of data, non-Markovian systems are non-trivial and their description requires the memory knowledge of past states. Here we analyse the feature of such memory by taking a simple 1D Heisenberg model as many-body Hamiltonian, and construct a non-Markovian description by representing the system over the single-particle reduced density matrix. The number of past states required for this representation to reproduce the time-dependent dynamics is found to grow exponentially with the number of spins and with the density of the system spectrum. Most importantly, we demonstrate that neural networks can work as time propagators at any time in the future and that they can be concatenated in time forming an autoregression. Such neural-network autoregression can be used to generate long-time and arbitrary dense time trajectories. Finally, we investigate the time resolution needed to represent the system memory. We find two regimes: for fine memory samplings the memory needed remains constant, while longer memories are required for coarse samplings, although the total number of time steps remains constant. The boundary between these two regimes is set by the period corresponding to the highest frequency in the system spectrum, demonstrating that neural network can overcome the limitation set by the Shannon-Nyquist sampling theorem.},
+ keywords = {_tablet,/unread,Condensed Matter - Mesoscale and Nanoscale Physics,Condensed Matter - Strongly Correlated Electrons,Quantum Physics},
+ file = {/Users/wasmer/Nextcloud/Zotero/Nelson et al_2022_Data-Driven Time Propagation of Quantum Systems with Neural Networks.pdf;/Users/wasmer/Zotero/storage/N33SL7SM/2201.html}
+}
+
+@article{nelsonMachineLearningDensity2019,
+ title = {Machine Learning Density Functional Theory for the {{Hubbard}} Model},
+ author = {Nelson, James and Tiwari, Rajarshi and Sanvito, Stefano},
+ date = {2019-02-15},
+ journaltitle = {Physical Review B},
+ shortjournal = {Phys. Rev. B},
+ volume = {99},
+ number = {7},
+ pages = {075132},
+ publisher = {{American Physical Society}},
+ doi = {10.1103/PhysRevB.99.075132},
+ url = {https://link.aps.org/doi/10.1103/PhysRevB.99.075132},
+ urldate = {2023-03-09},
+ abstract = {The solution of complex many-body lattice models can often be found by defining an energy functional of the relevant density of the problem. For instance, in the case of the Hubbard model the spin-resolved site occupation is enough to describe the system's total energy. Similarly to standard density functional theory, however, the exact functional is unknown, and suitable approximations need to be formulated. By using a deep-learning neural network trained on exact-diagonalization results, we demonstrate that one can construct an exact functional for the Hubbard model. In particular, we show that the neural network returns a ground-state energy numerically indistinguishable from that obtained by exact diagonalization and, most importantly, that the functional satisfies the two Hohenberg-Kohn theorems: for a given ground-state density it yields the external potential, and it is fully variational in the site occupation.},
+ keywords = {/unread,exact diagonaiization,Hubbard model,ML,ML-DFA,ML-DFT,ML-ESM,NN,prediction from density,prediction of ground-state properties,prediction of total energy,spin-dependent},
+ file = {/Users/wasmer/Nextcloud/Zotero/Nelson et al_2019_Machine learning density functional theory for the Hubbard model.pdf;/Users/wasmer/Zotero/storage/NRLX8AEL/PhysRevB.99.html}
+}
+
+@article{nelsonMachinelearningSemilocalDensity2021,
+ title = {Machine-Learning Semi-Local Density Functional Theory for Many-Body Lattice Models at Zero and Finite Temperature},
+ author = {Nelson, James and Tiwari, Rajarshi and Sanvito, Stefano},
+ date = {2021-06-07},
+ journaltitle = {Physical Review B},
+ shortjournal = {Phys. Rev. B},
+ volume = {103},
+ number = {24},
+ eprint = {2103.05510},
+ eprinttype = {arxiv},
+ eprintclass = {cond-mat, physics:quant-ph},
+ pages = {245111},
+ issn = {2469-9950, 2469-9969},
+ doi = {10.1103/PhysRevB.103.245111},
+ url = {http://arxiv.org/abs/2103.05510},
+ urldate = {2023-02-23},
+ abstract = {We introduce a machine-learning density-functional-theory formalism for the spinless Hubbard model in one dimension at both zero and finite temperature. In the zero-temperature case this establishes a one-to-one relation between the site occupation and the total energy, which is then minimised at the ground-state occupation. In contrast, at finite temperature the same relation is defined between the Helmholtz free energy and the equilibrium site occupation. Most importantly, both functionals are semi-local, so that they are independent from the size of the system under investigation and can be constructed over exact data for small systems. These 'exact' functionals are numerically defined by neural networks. We also define additional neural networks for finite-temperature thermodynamical quantities, such as the entropy and heat capacity. These can be either a functional of the ground-state site occupation or of the finite-temperature equilibrium site occupation. In the first case their equilibrium value does not correspond to an extremal point of the functional, while it does in the second case. Our work gives us access to finite-temperature properties of many-body systems in the thermodynamic limit.},
+ keywords = {_tablet,exact diagonaiization,finite-temperature DFT,Hubbard model,lattice DFT,ML,ML-DFA,ML-DFT,NN,thermodynamics},
+ file = {/Users/wasmer/Nextcloud/Zotero/Nelson et al_2021_Machine-learning semi-local density functional theory for many-body lattice.pdf;/Users/wasmer/Zotero/storage/FYR6QCEJ/2103.html}
+}
+
+@article{nelsonPredictingCurieTemperature2019,
+ title = {Predicting the {{Curie}} Temperature of Ferromagnets Using Machine Learning},
+ author = {Nelson, James and Sanvito, Stefano},
+ date = {2019-10-10},
+ journaltitle = {Physical Review Materials},
+ shortjournal = {Phys. Rev. Materials},
+ volume = {3},
+ number = {10},
+ eprint = {1906.08534},
+ eprinttype = {arxiv},
+ eprintclass = {cond-mat, physics:physics},
+ pages = {104405},
+ issn = {2475-9953},
+ doi = {10.1103/PhysRevMaterials.3.104405},
+ url = {http://arxiv.org/abs/1906.08534},
+ urldate = {2023-02-23},
+ abstract = {The magnetic properties of a material are determined by a subtle balance between the various interactions at play, a fact that makes the design of new magnets a daunting task. High-throughput electronic structure theory may help to explore the vast chemical space available and offers a design tool to the experimental synthesis. This method efficiently predicts the elementary magnetic properties of a compound and its thermodynamical stability, but it is blind to information concerning the magnetic critical temperature. Here we introduce a range of machine-learning models to predict the Curie temperature, \$T\_\textbackslash mathrm\{C\}\$, of ferromagnets. The models are constructed by using experimental data for about 2,500 known magnets and consider the chemical composition of a compound as the only feature determining \$T\_\textbackslash mathrm\{C\}\$. Thus, we are able to establish a one-to-one relation between the chemical composition and the critical temperature. We show that the best model can predict \$T\_\textbackslash mathrm\{C\}\$'s with an accuracy of about 50K. Most importantly our model is able to extrapolate the predictions to regions of the chemical space, where only a little fraction of the data was considered for training. This is demonstrated by tracing the \$T\_\textbackslash mathrm\{C\}\$ of binary intermetallic alloys along their composition space and for the Al-Co-Fe ternary system.},
+ keywords = {_tablet,/unread,Condensed Matter - Materials Science,Physics - Computational Physics},
+ file = {/Users/wasmer/Nextcloud/Zotero/Nelson_Sanvito_2019_Predicting the Curie temperature of ferromagnets using machine learning.pdf;/Users/wasmer/Zotero/storage/J4ASXLIA/1906.html}
}
@article{nemnesFeatureSelectionProcedures2021,
@@ -5277,7 +5830,23 @@
abstract = {This is an introductory machine learning course specifically developed with STEM students in mind. We discuss supervised, unsupervised, and reinforcement learning. The notes start with an exposition of machine learning methods without neural networks, such as principle component analysis, t-SNE, and linear regression. We continue with an introduction to both basic and advanced neural network structures such as conventional neural networks, (variational) autoencoders, generative adversarial networks, restricted Boltzmann machines, and recurrent neural networks. Questions of interpretability are discussed using the examples of dreaming and adversarial attacks.},
langid = {english},
keywords = {general,ML,review},
- file = {/home/johannes/Nextcloud/Zotero/Neupert et al_2021_Introduction to Machine Learning for the Sciences.pdf;/home/johannes/Zotero/storage/GE7KJ34Q/2102.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Neupert et al_2021_Introduction to Machine Learning for the Sciences.pdf;/Users/wasmer/Zotero/storage/GE7KJ34Q/2102.html}
+}
+
+@online{nigamCompletenessAtomicStructure2023,
+ title = {Completeness of {{Atomic Structure Representations}}},
+ author = {Nigam, Jigyasa and Pozdnyakov, Sergey N. and Huguenin-Dumittan, Kevin K. and Ceriotti, Michele},
+ date = {2023-02-28},
+ number = {arXiv:2302.14770},
+ eprint = {arXiv:2302.14770},
+ eprinttype = {arxiv},
+ doi = {10.48550/arXiv.2302.14770},
+ url = {http://arxiv.org/abs/2302.14770},
+ urldate = {2023-03-01},
+ abstract = {Achieving a complete and symmetric description of a group of point particles, such as atoms in a molecule, is a common problem in physics and theoretical chemistry. The introduction of machine learning to science has made this issue even more critical, as it underpins the ability of a model to reproduce arbitrary physical relationships, and to do so while being consistent with basic symmetries and conservation laws. However, the descriptors that are commonly used to represent point clouds -- most notably those adopted to describe matter at the atomic scale -- are unable to distinguish between special arrangements of particles. This makes it impossible to machine learn their properties. Frameworks that are provably complete exist, but are only so in the limit in which they simultaneously describe the mutual relationship between all atoms, which is impractical. We introduce, and demonstrate on a particularly insidious class of atomic arrangements, a strategy to build descriptors that rely solely on information on the relative arrangement of triplets of particles, but can be used to construct symmetry-adapted models that have universal approximation power.},
+ pubstate = {preprint},
+ keywords = {/unread,ACDC,descriptors,GNN,incompleteness,representation learning},
+ file = {/Users/wasmer/Nextcloud/Zotero/Nigam et al_2023_Completeness of Atomic Structure Representations.pdf;/Users/wasmer/Zotero/storage/T3WEIUD6/2302.html}
}
@article{nigamEquivariantRepresentationsMolecular2022,
@@ -5296,7 +5865,7 @@
urldate = {2022-10-04},
abstract = {Symmetry considerations are at the core of the major frameworks used to provide an effective mathematical representation of atomic configurations that is then used in machine-learning models to predict the properties associated with each structure. In most cases, the models rely on a description of atom-centered environments and are suitable to learn atomic properties or global observables that can be decomposed into atomic contributions. Many quantities that are relevant for quantum mechanical calculations, however—most notably the single-particle Hamiltonian matrix when written in an atomic orbital basis—are not associated with a single center, but with two (or more) atoms in the structure. We discuss a family of structural descriptors that generalize the very successful atom-centered density correlation features to the N-center case and show, in particular, how this construction can be applied to efficiently learn the matrix elements of the (effective) single-particle Hamiltonian written in an atom-centered orbital basis. These N-center features are fully equivariant—not only in terms of translations and rotations but also in terms of permutations of the indices associated with the atoms—and are suitable to construct symmetry-adapted machine-learning models of new classes of properties of molecules and materials.},
keywords = {ACDC,equivariant,ML,ML-DFT,ML-ESM,N-center representation,NICE,prediction of Hamiltonian matrix,with-code},
- file = {/home/johannes/Nextcloud/Zotero/Nigam et al_2022_Equivariant representations for molecular Hamiltonians and N-center.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Nigam et al_2022_Equivariant representations for molecular Hamiltonians and N-center.pdf}
}
@unpublished{nigamUnifiedTheoryAtomcentered2022,
@@ -5305,13 +5874,33 @@
date = {2022-02-03},
eprint = {2202.01566},
eprinttype = {arxiv},
- primaryclass = {physics, stat},
+ eprintclass = {physics, stat},
url = {http://arxiv.org/abs/2202.01566},
urldate = {2022-02-04},
abstract = {Data-driven schemes that associate molecular and crystal structures with their microscopic properties share the need for a concise, effective description of the arrangement of their atomic constituents. Many types of models rely on descriptions of atom-centered environments, that are associated with an atomic property or with an atomic contribution to an extensive macroscopic quantity. Frameworks in this class can be understood in terms of atom-centered density correlations (ACDC), that are used as a basis for a body-ordered, symmetry-adapted expansion of the targets. Several other schemes, that gather information on the relationship between neighboring atoms using graph-convolutional (or message-passing) ideas, cannot be directly mapped to correlations centered around a single atom. We generalize the ACDC framework to include multi-centered information, generating representations that provide a complete linear basis to regress symmetric functions of atomic coordinates, and form the basis to systematize our understanding of both atom-centered and graph-convolutional machine-learning schemes.},
- archiveprefix = {arXiv},
keywords = {_tablet,ACDC,ACE,descriptors,GCN,GNN,ML,MPNN,NequIP,NN,representation learning,SOAP,unified theory},
- file = {/home/johannes/Nextcloud/Zotero/false;/home/johannes/Nextcloud/Zotero/Nigam et al_2022_Unified theory of atom-centered representations and graph convolutional.pdf}
+ file = {/home/johannes/Nextcloud/Zotero/false;/Users/wasmer/Nextcloud/Zotero/Nigam et al_2022_Unified theory of atom-centered representations and graph convolutional.pdf}
+}
+
+@article{novikovMagneticMomentTensor2022,
+ title = {Magnetic {{Moment Tensor Potentials}} for Collinear Spin-Polarized Materials Reproduce Different Magnetic States of Bcc {{Fe}}},
+ author = {Novikov, Ivan and Grabowski, Blazej and Körmann, Fritz and Shapeev, Alexander},
+ date = {2022-01-25},
+ journaltitle = {npj Computational Materials},
+ shortjournal = {npj Comput Mater},
+ volume = {8},
+ number = {1},
+ pages = {1--6},
+ publisher = {{Nature Publishing Group}},
+ issn = {2057-3960},
+ doi = {10.1038/s41524-022-00696-9},
+ url = {https://www.nature.com/articles/s41524-022-00696-9},
+ urldate = {2023-02-21},
+ abstract = {We present the magnetic Moment Tensor Potentials (mMTPs), a class of machine-learning interatomic potentials, accurately reproducing both vibrational and magnetic degrees of freedom as provided, e.g., from first-principles calculations. The accuracy is achieved by a two-step minimization scheme that coarse-grains the atomic and the spin space. The performance of the mMTPs is demonstrated for the prototype magnetic system bcc iron, with applications to phonon calculations for different magnetic states, and molecular-dynamics simulations with fluctuating magnetic moments.},
+ issue = {1},
+ langid = {english},
+ keywords = {/unread,collinear,descriptors,magnetic moment,magnetism,MD,ML,MTP,phonon,spin-dependent,vectorial learning target},
+ file = {/Users/wasmer/Nextcloud/Zotero/Novikov et al_2022_Magnetic Moment Tensor Potentials for collinear spin-polarized materials.pdf}
}
@article{ohCompleteQuantumHall2013,
@@ -5327,13 +5916,12 @@
url = {https://www.science.org/doi/10.1126/science.1237215},
urldate = {2022-05-13},
keywords = {_tablet,Hall effect,Hall QAHE,Hall QHE,Hall QSHE,perspective},
- file = {/home/johannes/Nextcloud/Zotero/Oh_2013_The Complete Quantum Hall Trio.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Oh_2013_The Complete Quantum Hall Trio.pdf}
}
@article{oliveiraCECAMElectronicStructure2020,
title = {The {{CECAM}} Electronic Structure Library and the Modular Software Development Paradigm},
- author = {Oliveira, Micael J. T. and Papior, Nick and Pouillon, Yann and Blum, Volker and Artacho, Emilio and Caliste, Damien and Corsetti, Fabiano and de Gironcoli, Stefano and Elena, Alin M. and GarcÃa, Alberto and GarcÃa-Suárez, VÃctor M. and Genovese, Luigi and Huhn, William P. and Huhs, Georg and Kokott, Sebastian and Küçükbenli, Emine and Larsen, Ask H. and Lazzaro, Alfio and Lebedeva, Irina V. and Li, Yingzhou and López-Durán, David and López-Tarifa, Pablo and Lüders, Martin and Marques, Miguel A. L. and Minar, Jan and Mohr, Stephan and Mostofi, Arash A. and O’Cais, Alan and Payne, Mike C. and Ruh, Thomas and Smith, Daniel G. A. and Soler, José M. and Strubbe, David A. and Tancogne-Dejean, Nicolas and Tildesley, Dominic and Torrent, Marc and Yu, Victor Wen-zhe},
- options = {useprefix=true},
+ author = {Oliveira, Micael J. T. and Papior, Nick and Pouillon, Yann and Blum, Volker and Artacho, Emilio and Caliste, Damien and Corsetti, Fabiano and family=Gironcoli, given=Stefano, prefix=de, useprefix=true and Elena, Alin M. and GarcÃa, Alberto and GarcÃa-Suárez, VÃctor M. and Genovese, Luigi and Huhn, William P. and Huhs, Georg and Kokott, Sebastian and Küçükbenli, Emine and Larsen, Ask H. and Lazzaro, Alfio and Lebedeva, Irina V. and Li, Yingzhou and López-Durán, David and López-Tarifa, Pablo and Lüders, Martin and Marques, Miguel A. L. and Minar, Jan and Mohr, Stephan and Mostofi, Arash A. and O’Cais, Alan and Payne, Mike C. and Ruh, Thomas and Smith, Daniel G. A. and Soler, José M. and Strubbe, David A. and Tancogne-Dejean, Nicolas and Tildesley, Dominic and Torrent, Marc and Yu, Victor Wen-zhe},
date = {2020-07-14},
journaltitle = {The Journal of Chemical Physics},
shortjournal = {J. Chem. Phys.},
@@ -5346,7 +5934,7 @@
url = {https://aip.scitation.org/doi/10.1063/5.0012901},
urldate = {2021-10-26},
abstract = {First-principles electronic structure calculations are now accessible to a very large community of users across many disciplines, thanks to many successful software packages, some of which are described in this special issue. The traditional coding paradigm for such packages is monolithic, i.e., regardless of how modular its internal structure may be, the code is built independently from others, essentially from the compiler up, possibly with the exception of linear-algebra and message-passing libraries. This model has endured and been quite successful for decades. The successful evolution of the electronic structure methodology itself, however, has resulted in an increasing complexity and an ever longer list of features expected within all software packages, which implies a growing amount of replication between different packages, not only in the initial coding but, more importantly, every time a code needs to be re-engineered to adapt to the evolution of computer hardware architecture. The Electronic Structure Library (ESL) was initiated by CECAM (the European Centre for Atomic and Molecular Calculations) to catalyze a paradigm shift away from the monolithic model and promote modularization, with the ambition to extract common tasks from electronic structure codes and redesign them as open-source libraries available to everybody. Such libraries include “heavy-duty†ones that have the potential for a high degree of parallelization and adaptation to novel hardware within them, thereby separating the sophisticated computer science aspects of performance optimization and re-engineering from the computational science done by, e.g., physicists and chemists when implementing new ideas. We envisage that this modular paradigm will improve overall coding efficiency and enable specialists (whether they be computer scientists or computational scientists) to use their skills more effectively and will lead to a more dynamic evolution of software in the community as well as lower barriers to entry for new developers. The model comes with new challenges, though. The building and compilation of a code based on many interdependent libraries (and their versions) is a much more complex task than that of a code delivered in a single self-contained package. Here, we describe the state of the ESL, the different libraries it now contains, the short- and mid-term plans for further libraries, and the way the new challenges are faced. The ESL is a community initiative into which several pre-existing codes and their developers have contributed with their software and efforts, from which several codes are already benefiting, and which remains open to the community.},
- file = {/home/johannes/Nextcloud/Zotero/Oliveira et al_2020_The CECAM electronic structure library and the modular software development.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Oliveira et al_2020_The CECAM electronic structure library and the modular software development.pdf}
}
@article{onatSensitivityDimensionalityAtomic2020,
@@ -5365,7 +5953,7 @@
urldate = {2021-05-13},
abstract = {Faithfully representing chemical environments is essential for describing materials and molecules with machine learning approaches. Here, we present a systematic classification of these representations and then investigate (i) the sensitivity to perturbations and (ii) the effective dimensionality of a variety of atomic environment representations and over a range of material datasets. Representations investigated include atom centered symmetry functions, Chebyshev Polynomial Symmetry Functions (CHSF), smooth overlap of atomic positions, many-body tensor representation, and atomic cluster expansion. In area (i), we show that none of the atomic environment representations are linearly stable under tangential perturbations and that for CHSF, there are instabilities for particular choices of perturbation, which we show can be removed with a slight redefinition of the representation. In area (ii), we find that most representations can be compressed significantly without loss of precision and, further, that selecting optimal subsets of a representation method improves the accuracy of regression models built for a given dataset.},
keywords = {_tablet,ACE,ACSF,descriptors,MBTR,ML,MTP,SNAP,SOAP},
- file = {/home/johannes/Nextcloud/Zotero/Onat et al_2020_Sensitivity and dimensionality of atomic environment representations used for.pdf;/home/johannes/Zotero/storage/RQ8UAKFX/5.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Onat et al_2020_Sensitivity and dimensionality of atomic environment representations used for.pdf;/Users/wasmer/Zotero/storage/RQ8UAKFX/5.html}
}
@online{OnlineCourseMachine,
@@ -5375,7 +5963,7 @@
urldate = {2021-05-13},
abstract = {\# Online Course: Machine Learning for Physicists 2021 :::info **Lecture Series by Florian Marquard},
keywords = {course,ML,notebook,with-code},
- file = {/home/johannes/Zotero/storage/6TZCQAXX/Machine_Learning_For_Physicists_2021.html}
+ file = {/Users/wasmer/Zotero/storage/6TZCQAXX/Machine_Learning_For_Physicists_2021.html}
}
@article{ouyangSISSOCompressedsensingMethod2018,
@@ -5394,7 +5982,7 @@
urldate = {2021-05-19},
abstract = {The lack of reliable methods for identifying descriptors—the sets of parameters capturing the underlying mechanisms of a material's property—is one of the key factors hindering efficient materials development. Here, we propose a systematic approach for discovering descriptors for materials' properties, within the framework of compressed-sensing-based dimensionality reduction. The sure independence screening and sparsifying operator (SISSO) tackles immense and correlated features spaces, and converges to the optimal solution from a combination of features relevant to the materials' property of interest. In addition, SISSO gives stable results also with small training sets. The methodology is benchmarked with the quantitative prediction of the ground-state enthalpies of octet binary materials (using ab initio data) and applied to the showcase example of predicting the metal/insulator classification of binaries (with experimental data). Accurate, predictive models are found in both cases. For the metal-insulator classification model, the predictive capability is tested beyond the training data: It rediscovers the available pressure-induced insulator-to-metal transitions and it allows for the prediction of yet unknown transition candidates, ripe for experimental validation. As a step forward with respect to previous model-identification methods, SISSO can become an effective tool for automatic materials development.},
keywords = {compressed sensing,descriptors,descriptors analysis,ML},
- file = {/home/johannes/Nextcloud/Zotero/Ouyang et al_2018_SISSO.pdf;/home/johannes/Zotero/storage/FPEWTJ64/PhysRevMaterials.2.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Ouyang et al_2018_SISSO.pdf;/Users/wasmer/Zotero/storage/FPEWTJ64/PhysRevMaterials.2.html}
}
@article{oviedoInterpretableExplainableMachine2022,
@@ -5412,7 +6000,7 @@
urldate = {2022-07-11},
abstract = {ConspectusMachine learning has become a common and powerful tool in materials research. As more data become available, with the use of high-performance computing and high-throughput experimentation, machine learning has proven potential to accelerate scientific research and technology development. Though the uptake of data-driven approaches for materials science is at an exciting, early stage, to realize the true potential of machine learning models for successful scientific discovery, they must have qualities beyond purely predictive power. The predictions and inner workings of models should provide a certain degree of explainability by human experts, permitting the identification of potential model issues or limitations, building trust in model predictions, and unveiling unexpected correlations that may lead to scientific insights. In this work, we summarize applications of interpretability and explainability techniques for materials science and chemistry and discuss how these techniques can improve the outcome of scientific studies. We start by defining the fundamental concepts of interpretability and explainability in machine learning and making them less abstract by providing examples in the field. We show how interpretability in scientific machine learning has additional constraints compared to general applications. Building upon formal definitions in machine learning, we formulate the basic trade-offs among the explainability, completeness, and scientific validity of model explanations in scientific problems. In the context of these trade-offs, we discuss how interpretable models can be constructed, what insights they provide, and what drawbacks they have. We present numerous examples of the application of interpretable machine learning in a variety of experimental and simulation studies, encompassing first-principles calculations, physicochemical characterization, materials development, and integration into complex systems. We discuss the varied impacts and uses of interpretabiltiy in these cases according to the nature and constraints of the scientific study of interest. We discuss various challenges for interpretable machine learning in materials science and, more broadly, in scientific settings. In particular, we emphasize the risks of inferring causation or reaching generalization by purely interpreting machine learning models and the need for uncertainty estimates for model explanations. Finally, we showcase a number of exciting developments in other fields that could benefit interpretability in material science problems. Adding interpretability to a machine learning model often requires no more technical know-how than building the model itself. By providing concrete examples of studies (many with associated open source code and data), we hope that this Account will encourage all practitioners of machine learning in materials science to look deeper into their models.},
keywords = {AML,ML,XAI},
- file = {/home/johannes/Nextcloud/Zotero/Oviedo et al_2022_Interpretable and Explainable Machine Learning for Materials Science and.pdf;/home/johannes/Zotero/storage/9I3JM9FX/accountsmr.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Oviedo et al_2022_Interpretable and Explainable Machine Learning for Materials Science and.pdf;/Users/wasmer/Zotero/storage/9I3JM9FX/accountsmr.html}
}
@article{paleicoBinHashMethod2021,
@@ -5433,12 +6021,12 @@
langid = {english},
keywords = {ACSF,descriptors,descriptors analysis,ML,MLP},
annotation = {0 citations (Crossref) [2021-05-04]},
- file = {/home/johannes/Nextcloud/Zotero/Paleico_Behler_2021_A bin and hash method for analyzing reference data and descriptors in machine.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Paleico_Behler_2021_A bin and hash method for analyzing reference data and descriptors in machine.pdf}
}
@article{parsaeifardAssessmentStructuralResolution2021,
title = {An Assessment of the Structural Resolution of Various Fingerprints Commonly Used in Machine Learning},
- author = {Parsaeifard, Behnam and De, Deb Sankar and Christensen, Anders S. and Faber, Felix A. and Kocer, Emir and De, Sandip and Behler, Jörg and von Lilienfeld, O. Anatole and Goedecker, Stefan},
+ author = {Parsaeifard, Behnam and De, Deb Sankar and Christensen, Anders S. and Faber, Felix A. and Kocer, Emir and De, Sandip and Behler, Jörg and family=Lilienfeld, given=O. Anatole, prefix=von, useprefix=false and Goedecker, Stefan},
date = {2021-03},
journaltitle = {Machine Learning: Science and Technology},
shortjournal = {Mach. Learn.: Sci. Technol.},
@@ -5453,7 +6041,7 @@
abstract = {Atomic environment fingerprints are widely used in computational materials science, from machine learning potentials to the quantification of similarities between atomic configurations. Many approaches to the construction of such fingerprints, also called structural descriptors, have been proposed. In this work, we compare the performance of fingerprints based on the overlap matrix, the smooth overlap of atomic positions, Behler–Parrinello atom-centered symmetry functions, modified Behler–Parrinello symmetry functions used in the ANI-1ccx potential and the Faber–Christensen–Huang–Lilienfeld fingerprint under various aspects. We study their ability to resolve differences in local environments and in particular examine whether there are certain atomic movements that leave the fingerprints exactly or nearly invariant. For this purpose, we introduce a sensitivity matrix whose eigenvalues quantify the effect of atomic displacement modes on the fingerprint. Further, we check whether these displacements correlate with the variation of localized physical quantities such as forces. Finally, we extend our examination to the correlation between molecular fingerprints obtained from the atomic fingerprints and global quantities of entire molecules.},
langid = {english},
keywords = {ACSF,BPSF,descriptor comparison,descriptors,FCHL,ML,SOAP},
- file = {/home/johannes/Nextcloud/Zotero/Parsaeifard et al_2021_An assessment of the structural resolution of various fingerprints commonly.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Parsaeifard et al_2021_An assessment of the structural resolution of various fingerprints commonly.pdf}
}
@article{parsaeifardManifoldsQuasiconstantSOAP2022,
@@ -5472,7 +6060,7 @@
urldate = {2022-09-20},
abstract = {Atomic fingerprints are commonly used for the characterization of local environments of atoms in machine learning and other contexts. In this work, we study the behavior of two widely used fingerprints, namely, the smooth overlap of atomic positions (SOAP) and the atom-centered symmetry functions (ACSFs), under finite changes of atomic positions and demonstrate the existence of manifolds of quasi-constant fingerprints. These manifolds are found numerically by following eigenvectors of the sensitivity matrix with quasi-zero eigenvalues. The existence of such manifolds in ACSF and SOAP causes a failure to machine learn four-body interactions, such as torsional energies that are part of standard force fields. No such manifolds can be found for the overlap matrix (OM) fingerprint due to its intrinsic many-body character.},
keywords = {ACSF,descriptors,descriptors analysis,incompleteness,ML,SOAP},
- file = {/home/johannes/Nextcloud/Zotero/Parsaeifard_Goedecker_2022_Manifolds of quasi-constant SOAP and ACSF fingerprints and the resulting.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Parsaeifard_Goedecker_2022_Manifolds of quasi-constant SOAP and ACSF fingerprints and the resulting.pdf}
}
@unpublished{pedersonMachineLearningDensity2022,
@@ -5481,13 +6069,12 @@
date = {2022-05-03},
eprint = {2205.01591},
eprinttype = {arxiv},
- primaryclass = {cond-mat, physics:physics},
+ eprintclass = {cond-mat, physics:physics},
url = {http://arxiv.org/abs/2205.01591},
urldate = {2022-05-13},
abstract = {Over the past decade machine learning has made significant advances in approximating density functionals, but whether this signals the end of human-designed functionals remains to be seen. Ryan Pederson, Bhupalee Kalita and Kieron Burke discuss the rise of machine learning for functional design.},
- archiveprefix = {arXiv},
keywords = {DeepMind,density functional,DFT,DM21,ML,ML-DFT,ML-ESM},
- file = {/home/johannes/Nextcloud/Zotero/Pederson et al_2022_Machine learning and density functional theory.pdf;/home/johannes/Zotero/storage/UPT9RJEW/2205.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Pederson et al_2022_Machine learning and density functional theory.pdf;/Users/wasmer/Zotero/storage/UPT9RJEW/2205.html}
}
@unpublished{pedregosaScikitlearnMachineLearning2018,
@@ -5497,13 +6084,12 @@
date = {2018-06-05},
eprint = {1201.0490},
eprinttype = {arxiv},
- primaryclass = {cs},
+ eprintclass = {cs},
url = {http://arxiv.org/abs/1201.0490},
urldate = {2021-07-14},
abstract = {Scikit-learn is a Python module integrating a wide range of state-of-the-art machine learning algorithms for medium-scale supervised and unsupervised problems. This package focuses on bringing machine learning to non-specialists using a general-purpose high-level language. Emphasis is put on ease of use, performance, documentation, and API consistency. It has minimal dependencies and is distributed under the simplified BSD license, encouraging its use in both academic and commercial settings. Source code, binaries, and documentation can be downloaded from http://scikit-learn.org.},
- archiveprefix = {arXiv},
keywords = {library,ML,scikit-learn},
- file = {/home/johannes/Nextcloud/Zotero/Pedregosa et al_2018_Scikit-learn.pdf;/home/johannes/Zotero/storage/RDIVNKI6/1201.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Pedregosa et al_2018_Scikit-learn.pdf;/Users/wasmer/Zotero/storage/RDIVNKI6/1201.html}
}
@article{peixotoNonlocalEffectImpurity2020,
@@ -5524,7 +6110,7 @@
issue = {1},
langid = {english},
keywords = {Electronic properties and materials,Ferromagnetism,Hall QAHE,Magnetic properties and materials,ruess,Spintronics,topological insulator,Topological insulators},
- file = {/home/johannes/Nextcloud/Zotero/Peixoto et al_2020_Non-local effect of impurity states on the exchange coupling mechanism in.pdf;/home/johannes/Zotero/storage/DDIQNTSB/s41535-020-00288-0.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Peixoto et al_2020_Non-local effect of impurity states on the exchange coupling mechanism in.pdf;/Users/wasmer/Zotero/storage/DDIQNTSB/s41535-020-00288-0.html}
}
@article{pereiraChallengesTopologicalInsulator2021,
@@ -5543,8 +6129,7 @@
abstract = {Topological insulators (TIs) are of particular interest in the recent solid-state research because of their exceptional features stemming from the conducting, topologically protected surface states. The exotic properties include the occurrence of novel quantum phenomena and make them promising materials for spintronics and quantum computing applications. Theoretical studies have provided a vast amount of valuable predictions and proposals, whose experimental observation and implementation, to date, are often hindered by an insufficient sample quality. The effect of even a relatively low concentration of defects can make the access to purely topological surface states impossible. This points out the need of high-quality bulk-insulating materials with ultra-clean surfaces/interfaces, which requires sophisticated sample/device preparations as well as special precautions during the measurements. Herein, the challenging work on 3D TI thin films with a focus on Bi2Te3 is reported. It covers the optimization of the molecular beam epitaxy growth process, the in situ characterization of surface states and transport properties, the influence of exposure to ambient gases and of capping layers, as well as the effect of interfacing TI thin film with magnetic materials.},
langid = {english},
keywords = {angle-resolved photoelectron spectroscopy,in situ transport,molecular beam epitaxy,topological insulators},
- annotation = {\_eprint: https://onlinelibrary.wiley.com/doi/pdf/10.1002/pssb.202000346},
- file = {/home/johannes/Nextcloud/Zotero/Pereira et al_2021_Challenges of Topological Insulator Research.pdf;/home/johannes/Zotero/storage/CXAHTBAM/pssb.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Pereira et al_2021_Challenges of Topological Insulator Research.pdf;/Users/wasmer/Zotero/storage/CXAHTBAM/pssb.html}
}
@inproceedings{pezoaFoundationsJSONSchema2016,
@@ -5562,7 +6147,7 @@
abstract = {JSON -- the most popular data format for sending API requests and responses -- is still lacking a standardized schema or meta-data definition that allows the developers to specify the structure of JSON documents. JSON Schema is an attempt to provide a general purpose schema language for JSON, but it is still work in progress, and the formal specification has not yet been agreed upon. Why this could be a problem becomes evident when examining the behaviour of numerous tools for validating JSON documents against this initial schema proposal: although they agree on most general cases, when presented with the greyer areas of the specification they tend to differ significantly. In this paper we provide the first formal definition of syntax and semantics for JSON Schema and use it to show that implementing this layer on top of JSON is feasible in practice. This is done both by analysing the theoretical aspects of the validation problem and by showing how to set up and validate a JSON Schema for Wikidata, the central storage for Wikimedia.},
isbn = {978-1-4503-4143-1},
keywords = {expressiveness of schema languages,JSON,JSON schema,JSON validation},
- file = {/home/johannes/Nextcloud/Zotero/Pezoa et al_2016_Foundations of JSON Schema.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Pezoa et al_2016_Foundations of JSON Schema.pdf}
}
@article{pfauInitioSolutionManyelectron2020,
@@ -5575,7 +6160,7 @@
number = {3},
doi = {10.1103/PhysRevResearch.2.033429},
keywords = {DeepMind,FermiNet,JAX,library,MC,ML,ML-ESM,ML-QMBP,NN,PauliNet,prediction of wavefunction,QMC,VMC,with-code},
- file = {/home/johannes/Nextcloud/Zotero/Pfau_2020_iAb initio-i solution of the many-electron Schrödinger equation with deep.pdf;/home/johannes/Zotero/storage/7HFHVNYZ/PhysRevResearch.2.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Pfau_2020_iAb initio-i solution of the many-electron Schrödinger equation with deep.pdf;/Users/wasmer/Zotero/storage/7HFHVNYZ/PhysRevResearch.2.html}
}
@article{pilaniaDataBasedMethodsMaterials2020,
@@ -5593,7 +6178,27 @@
url = {https://www.morganclaypool.com/doi/10.2200/S00981ED1V01Y202001MOP001},
urldate = {2021-05-19},
keywords = {book,materials informatics},
- file = {/home/johannes/Nextcloud/Zotero/Pilania et al_2020_Data-Based Methods for Materials Design and Discovery.pdf;/home/johannes/Zotero/storage/8YQEF8LU/S00981ED1V01Y202001MOP001.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Pilania et al_2020_Data-Based Methods for Materials Design and Discovery.pdf;/Users/wasmer/Zotero/storage/8YQEF8LU/S00981ED1V01Y202001MOP001.html}
+}
+
+@article{pinheiroChoosingRightMolecular2021,
+ title = {Choosing the Right Molecular Machine Learning Potential},
+ author = {Pinheiro, Max and Ge, Fuchun and Ferré, Nicolas and Dral, Pavlo O. and Barbatti, Mario},
+ date = {2021-11-10},
+ journaltitle = {Chemical Science},
+ shortjournal = {Chem. Sci.},
+ volume = {12},
+ number = {43},
+ pages = {14396--14413},
+ publisher = {{The Royal Society of Chemistry}},
+ issn = {2041-6539},
+ doi = {10.1039/D1SC03564A},
+ url = {https://pubs.rsc.org/en/content/articlelanding/2021/sc/d1sc03564a},
+ urldate = {2023-03-02},
+ abstract = {Quantum-chemistry simulations based on potential energy surfaces of molecules provide invaluable insight into the physicochemical processes at the atomistic level and yield such important observables as reaction rates and spectra. Machine learning potentials promise to significantly reduce the computational cost and hence enable otherwise unfeasible simulations. However, the surging number of such potentials begs the question of which one to choose or whether we still need to develop yet another one. Here, we address this question by evaluating the performance of popular machine learning potentials in terms of accuracy and computational cost. In addition, we deliver structured information for non-specialists in machine learning to guide them through the maze of acronyms, recognize each potential's main features, and judge what they could expect from each one.},
+ langid = {english},
+ keywords = {/unread,AML,benchmarking,kernel methods,KRR,ML,MLatom,MLP,NN,with-code},
+ file = {/Users/wasmer/Nextcloud/Zotero/Pinheiro et al_2021_Choosing the right molecular machine learning potential.pdf}
}
@article{poelkingBenchMLExtensiblePipelining2022,
@@ -5610,7 +6215,7 @@
abstract = {We introduce a machine-learning (ML) framework for high-throughput benchmarking of diverse representations of chemical systems against datasets of materials and molecules. The guiding principle underlying the benchmarking approach is to evaluate raw descriptor performance by limiting model complexity to simple regression schemes while enforcing best ML practices, allowing for unbiased hyperparameter optimization, and assessing learning progress through learning curves along series of synchronized train-test splits. The resulting models are intended as baselines that can inform future method development, in addition to indicating how easily a given dataset can be learnt. Through a comparative analysis of the training outcome across a diverse set of physicochemical, topological and geometric representations, we glean insight into the relative merits of these representations as well as their interrelatedness.},
langid = {english},
keywords = {descriptor comparison},
- file = {/home/johannes/Nextcloud/Zotero/Poelking et al_2022_BenchML.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Poelking et al_2022_BenchML.pdf}
}
@online{PossibleGameChanger,
@@ -5619,7 +6224,7 @@
urldate = {2022-12-07},
langid = {english},
keywords = {/unread,condensed matter,physics,popular science,skyrmions,Spintronics},
- file = {/home/johannes/Zotero/storage/9GDZNCPD/a-possible-game-changer-for-next-generation-microelectronics.html}
+ file = {/Users/wasmer/Zotero/storage/9GDZNCPD/a-possible-game-changer-for-next-generation-microelectronics.html}
}
@article{pozdnyakovIncompletenessAtomicStructure2020,
@@ -5637,7 +6242,7 @@
urldate = {2021-05-13},
abstract = {Many-body descriptors are widely used to represent atomic environments in the construction of machine-learned interatomic potentials and more broadly for fitting, classification, and embedding tasks on atomic structures. There is a widespread belief in the community that three-body correlations are likely to provide an overcomplete description of the environment of an atom. We produce several counterexamples to this belief, with the consequence that any classifier, regression, or embedding model for atom-centered properties that uses three- (or four)-body features will incorrectly give identical results for different configurations. Writing global properties (such as total energies) as a sum of many atom-centered contributions mitigates the impact of this fundamental deficiency—explaining the success of current “machine-learning†force fields. We anticipate the issues that will arise as the desired accuracy increases, and suggest potential solutions.},
keywords = {_tablet,3-body order descriptors,descriptors,descriptors analysis,GPR,incompleteness,MBTR,ML,SOAP},
- file = {/home/johannes/Nextcloud/Zotero/Pozdnyakov et al_2020_Incompleteness of Atomic Structure Representations.pdf;/home/johannes/Zotero/storage/5QHMC4CR/PhysRevLett.125.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Pozdnyakov et al_2020_Incompleteness of Atomic Structure Representations.pdf;/Users/wasmer/Zotero/storage/5QHMC4CR/PhysRevLett.125.html}
}
@article{pozdnyakovIncompletenessGraphConvolutional2022,
@@ -5650,7 +6255,7 @@
abstract = {Graph neural networks (GNN) are very popular methods in machine learning and have been applied very successfully to the prediction of the properties of molecules and materials. First-order GNNs are well known to be incomplete, i.e., there exist graphs that are distinct but appear identical when seen through the lens of the GNN. More complicated schemes have thus been designed to increase their resolving power. Applications to molecules (and more generally, point clouds), however, add a geometric dimension to the problem. The most straightforward and prevalent approach to construct graph representation for molecules regards atoms as vertices in a graph and draws a bond between each pair of atoms within a chosen cutoff. Bonds can be decorated with the distance between atoms, and the resulting "distance graph NNs" (dGNN) have empirically demonstrated excellent resolving power and are widely used in chemical ML, with all known indistinguishable graphs being resolved in the fully-connected limit. Here we show that even for the restricted case of fully-connected graphs induced by 3D atom clouds dGNNs are not complete. We construct pairs of distinct point clouds that generate graphs that, for any cutoff radius, are equivalent based on a first-order Weisfeiler-Lehman test. This class of degenerate structures includes chemically-plausible configurations, setting an ultimate limit to the expressive power of some of the well-established GNN architectures for atomistic machine learning. Models that explicitly use angular or directional information in the description of atomic environments can resolve these degeneracies.},
langid = {english},
keywords = {GCN,GNN,incompleteness,ML,MPNN,WL test},
- file = {/home/johannes/Nextcloud/Zotero/Pozdnyakov_Ceriotti_2022_Incompleteness of graph convolutional neural networks for points clouds in.pdf;/home/johannes/Zotero/storage/ZKHDUH3X/2201.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Pozdnyakov_Ceriotti_2022_Incompleteness of graph convolutional neural networks for points clouds in.pdf;/Users/wasmer/Zotero/storage/ZKHDUH3X/2201.html}
}
@article{prodanNearsightednessElectronicMatter2005,
@@ -5666,7 +6271,23 @@
url = {https://www.pnas.org/doi/full/10.1073/pnas.0505436102},
urldate = {2022-10-05},
keywords = {condensed matter,electronic structure,near-sightedness,NEM,original publication,physics},
- file = {/home/johannes/Nextcloud/Zotero/Prodan_Kohn_2005_Nearsightedness of electronic matter.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Prodan_Kohn_2005_Nearsightedness of electronic matter.pdf}
+}
+
+@online{qamarAtomicClusterExpansion2022,
+ title = {Atomic Cluster Expansion for Quantum-Accurate Large-Scale Simulations of Carbon},
+ author = {Qamar, Minaam and Mrovec, Matous and Lysogorskiy, Yury and Bochkarev, Anton and Drautz, Ralf},
+ date = {2022-10-25},
+ number = {arXiv:2210.09161},
+ eprint = {arXiv:2210.09161},
+ eprinttype = {arxiv},
+ doi = {10.48550/arXiv.2210.09161},
+ url = {http://arxiv.org/abs/2210.09161},
+ urldate = {2023-01-20},
+ abstract = {We present an atomic cluster expansion (ACE) for carbon that improves over available classical and machine learning potentials. The ACE is parameterized from an exhaustive set of important carbon structures at extended volume and energy range, computed using density functional theory (DFT). Rigorous validation reveals that ACE predicts accurately a broad range of properties of both crystalline and amorphous carbon phases while being several orders of magnitude more computationally efficient than available machine learning models. We demonstrate the predictive power of ACE on three distinct applications, brittle crack propagation in diamond, evolution of amorphous carbon structures at different densities and quench rates and nucleation and growth of fullerene clusters under high pressure and temperature conditions.},
+ pubstate = {preprint},
+ keywords = {/unread,ACE,carbon,descriptors,MLP,molecular dynamics},
+ file = {/Users/wasmer/Nextcloud/Zotero/Qamar et al_2022_Atomic cluster expansion for quantum-accurate large-scale simulations of carbon.pdf;/Users/wasmer/Zotero/storage/SCVIRYIV/2210.html}
}
@book{QuantumTheoryMagnetism,
@@ -5675,7 +6296,7 @@
urldate = {2022-06-18},
langid = {english},
keywords = {_tablet,condensed matter,graduate,magnetism,textbook},
- file = {/home/johannes/Nextcloud/Zotero/Quantum Theory of Magnetism.pdf;/home/johannes/Zotero/storage/ULV44ULF/978-3-540-85416-6.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Quantum Theory of Magnetism.pdf;/Users/wasmer/Zotero/storage/ULV44ULF/978-3-540-85416-6.html}
}
@unpublished{rackersCrackingQuantumScaling2022,
@@ -5685,33 +6306,46 @@
number = {arXiv:2201.03726},
eprint = {2201.03726},
eprinttype = {arxiv},
- primaryclass = {cond-mat, physics:physics},
+ eprintclass = {cond-mat, physics:physics},
publisher = {{arXiv}},
doi = {10.48550/arXiv.2201.03726},
url = {http://arxiv.org/abs/2201.03726},
urldate = {2022-05-18},
abstract = {A long-standing goal of science is to accurately solve the Schr\textbackslash "odinger equation for large molecular systems. The poor scaling of current quantum chemistry algorithms on classical computers imposes an effective limit of about a few dozen atoms for which we can calculate molecular electronic structure. We present a machine learning (ML) method to break through this scaling limit and make quantum chemistry calculations of very large systems possible. We show that Euclidean Neural Networks can be trained to predict the electron density with high fidelity from limited data. Learning the electron density allows us to train a machine learning model on small systems and make accurate predictions on large ones. We show that this ML electron density model can break through the quantum scaling limit and calculate the electron density of systems of thousands of atoms with quantum accuracy.},
- archiveprefix = {arXiv},
keywords = {_tablet,charge density,e3nn,ENN,ML,ML-DFT,ML-ESM,molecules,prediction of electron density,script,target: density,transfer learning,with-code},
- file = {/home/johannes/Nextcloud/Zotero/Rackers et al_2022_Cracking the Quantum Scaling Limit with Machine Learned Electron Densities.pdf;/home/johannes/Zotero/storage/X9XGJLLI/2201.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Rackers et al_2022_Cracking the Quantum Scaling Limit with Machine Learned Electron Densities.pdf;/Users/wasmer/Zotero/storage/X9XGJLLI/2201.html}
}
-@misc{rackersCrackingQuantumScaling2022a,
+@online{rackersCrackingQuantumScaling2022a,
title = {Cracking the {{Quantum Scaling Limit}} with {{Machine Learned Electron Densities}}},
author = {Rackers, Joshua A. and Tecot, Lucas and Geiger, Mario and Smidt, Tess E.},
date = {2022-02-10},
number = {arXiv:2201.03726},
- eprint = {2201.03726},
+ eprint = {arXiv:2201.03726},
eprinttype = {arxiv},
- primaryclass = {cond-mat, physics:physics},
- publisher = {{arXiv}},
doi = {10.48550/arXiv.2201.03726},
url = {http://arxiv.org/abs/2201.03726},
urldate = {2022-07-10},
abstract = {A long-standing goal of science is to accurately solve the Schr\textbackslash "odinger equation for large molecular systems. The poor scaling of current quantum chemistry algorithms on classical computers imposes an effective limit of about a few dozen atoms for which we can calculate molecular electronic structure. We present a machine learning (ML) method to break through this scaling limit and make quantum chemistry calculations of very large systems possible. We show that Euclidean Neural Networks can be trained to predict the electron density with high fidelity from limited data. Learning the electron density allows us to train a machine learning model on small systems and make accurate predictions on large ones. We show that this ML electron density model can break through the quantum scaling limit and calculate the electron density of systems of thousands of atoms with quantum accuracy.},
- archiveprefix = {arXiv},
+ pubstate = {preprint},
keywords = {Condensed Matter - Soft Condensed Matter,Physics - Biological Physics,Physics - Chemical Physics},
- file = {/home/johannes/Nextcloud/Zotero/Rackers et al_2022_Cracking the Quantum Scaling Limit with Machine Learned Electron Densities2.pdf;/home/johannes/Zotero/storage/NL7QJTKF/2201.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Rackers et al_2022_Cracking the Quantum Scaling Limit with Machine Learned Electron Densities2.pdf;/Users/wasmer/Zotero/storage/NL7QJTKF/2201.html}
+}
+
+@article{rackersRecipeCrackingQuantum2023,
+ title = {A {{Recipe}} for {{Cracking}} the {{Quantum Scaling Limit}} with {{Machine Learned Electron Densities}}},
+ author = {Rackers, Joshua and Tecot, Lucas and Geiger, Mario and Smidt, Tess},
+ date = {2023},
+ journaltitle = {Machine Learning: Science and Technology},
+ shortjournal = {Mach. Learn.: Sci. Technol.},
+ issn = {2632-2153},
+ doi = {10.1088/2632-2153/acb314},
+ url = {http://iopscience.iop.org/article/10.1088/2632-2153/acb314},
+ urldate = {2023-01-20},
+ abstract = {A long-standing goal of science is to accurately simulate large molecular systems using quantum mechanics. The poor scaling of current quantum chemistry algorithms on classical computers, however, imposes an effective limit of about a few dozen atoms on traditional electronic structure calculations. We present a machine learning (ML) method to break through this scaling limit for electron densities. We show that Euclidean Neural Networks can be trained to predict molecular electron densities from limited data. By learning the electron density, the model can be trained on small systems and make accurate predictions on large ones. In the context of water clusters, we show that an ML model trained on clusters of just 12 molecules contains all the information needed to make accurate electron density predictions on cluster sizes of 50 or more, beyond the scaling limit of current quantum chemistry methods.},
+ langid = {english},
+ keywords = {CCSD(T),e3nn,equivariant,GNN,ML-DFT,molecules,prediction of electron density,with-code},
+ file = {/Users/wasmer/Nextcloud/Zotero/Rackers et al_2023_A Recipe for Cracking the Quantum Scaling Limit with Machine Learned Electron.pdf}
}
@article{raderTopologicalInsulatorsMaterials2021,
@@ -5728,8 +6362,7 @@
url = {https://onlinelibrary.wiley.com/doi/abs/10.1002/pssb.202170010},
urldate = {2022-05-30},
abstract = {Topological insulators are materials that are electrically insulating in the bulk but can conduct electricity due to topologically protected electronic edge or surface states. Since 2013, the German Research Foundation (DFG) has been supporting the Priority Program “Topological Insulators: Materials – Fundamental Properties – Devices†(SPP 1666) and in the time since, topological insulators developed from a mere curiosity to a material class that entered many fields of applied research. This Special Issue presents in 20 articles reports of the Priority Program reflecting its three areas of activity: (i) Understanding and improvement of existing topological insulator materials, regarding the size of the band gap and intrinsic doping levels, to enable room temperature applications, (ii) exploration of fundamental properties necessary for the development of device structures, and (iii) discovery of new materials to overcome deficits of current materials and explore new properties. See also the Guest Editorial (article number 2000594). The cover shows a Bi2Te3/MnBi2Te4 heterostructure, a ferromagnetic topological insulator (data by M. Albu, H. Groiss, S. Wimmer, G. Kothleitner, O. Caha, and J. MichaliÄka; artwork by E. D. L. Rienks and O. Rader).},
- annotation = {\_eprint: https://onlinelibrary.wiley.com/doi/pdf/10.1002/pssb.202170010},
- file = {/home/johannes/Nextcloud/Zotero/Rader et al_2021_Topological Insulators.pdf;/home/johannes/Zotero/storage/3TIJLTAF/pssb.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Rader et al_2021_Topological Insulators.pdf;/Users/wasmer/Zotero/storage/3TIJLTAF/pssb.html}
}
@article{raderTopologicalInsulatorsMaterials2021a,
@@ -5746,8 +6379,7 @@
url = {https://onlinelibrary.wiley.com/doi/abs/10.1002/pssb.202000594},
urldate = {2022-05-30},
langid = {english},
- annotation = {\_eprint: https://onlinelibrary.wiley.com/doi/pdf/10.1002/pssb.202000594},
- file = {/home/johannes/Nextcloud/Zotero/Rader et al_2021_Topological Insulators2.pdf;/home/johannes/Zotero/storage/CXY5KUXP/pssb.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Rader et al_2021_Topological Insulators2.pdf;/Users/wasmer/Zotero/storage/CXY5KUXP/pssb.html}
}
@article{raissiPhysicsinformedNeuralNetworks2019,
@@ -5766,14 +6398,13 @@
abstract = {We introduce physics-informed neural networks – neural networks that are trained to solve supervised learning tasks while respecting any given laws of physics described by general nonlinear partial differential equations. In this work, we present our developments in the context of solving two main classes of problems: data-driven solution and data-driven discovery of partial differential equations. Depending on the nature and arrangement of the available data, we devise two distinct types of algorithms, namely continuous time and discrete time models. The first type of models forms a new family of data-efficient spatio-temporal function approximators, while the latter type allows the use of arbitrarily accurate implicit Runge–Kutta time stepping schemes with unlimited number of stages. The effectiveness of the proposed framework is demonstrated through a collection of classical problems in fluids, quantum mechanics, reaction–diffusion systems, and the propagation of nonlinear shallow-water waves.},
langid = {english},
keywords = {Nonlinear dynamics,original publication,PINN,Python,rec-by-bluegel,Runge–Kutta methods,with-code},
- file = {/home/johannes/Nextcloud/Zotero/Raissi et al_2019_Physics-informed neural networks.pdf;/home/johannes/Zotero/storage/YR4ICSGZ/S0021999118307125.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Raissi et al_2019_Physics-informed neural networks.pdf;/Users/wasmer/Zotero/storage/YR4ICSGZ/S0021999118307125.html}
}
@article{ramakrishnanBigDataMeets2015,
title = {Big {{Data Meets Quantum Chemistry Approximations}}: {{The Δ-Machine Learning Approach}}},
shorttitle = {Big {{Data Meets Quantum Chemistry Approximations}}},
- author = {Ramakrishnan, Raghunathan and Dral, Pavlo O. and Rupp, Matthias and von Lilienfeld, O. Anatole},
- options = {useprefix=true},
+ author = {Ramakrishnan, Raghunathan and Dral, Pavlo O. and Rupp, Matthias and family=Lilienfeld, given=O. Anatole, prefix=von, useprefix=true},
date = {2015-05-12},
journaltitle = {Journal of Chemical Theory and Computation},
shortjournal = {J. Chem. Theory Comput.},
@@ -5787,7 +6418,7 @@
urldate = {2022-09-29},
abstract = {Chemically accurate and comprehensive studies of the virtual space of all possible molecules are severely limited by the computational cost of quantum chemistry. We introduce a composite strategy that adds machine learning corrections to computationally inexpensive approximate legacy quantum methods. After training, highly accurate predictions of enthalpies, free energies, entropies, and electron correlation energies are possible, for significantly larger molecular sets than used for training. For thermochemical properties of up to 16k isomers of C7H10O2 we present numerical evidence that chemical accuracy can be reached. We also predict electron correlation energy in post Hartree–Fock methods, at the computational cost of Hartree–Fock, and we establish a qualitative relationship between molecular entropy and electron correlation. The transferability of our approach is demonstrated, using semiempirical quantum chemistry and machine learning models trained on 1 and 10\% of 134k organic molecules, to reproduce enthalpies of all remaining molecules at density functional theory level of accuracy.},
keywords = {delta learning,Δ-machine learning},
- file = {/home/johannes/Nextcloud/Zotero/Ramakrishnan et al_2015_Big Data Meets Quantum Chemistry Approximations.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Ramakrishnan et al_2015_Big Data Meets Quantum Chemistry Approximations.pdf}
}
@unpublished{ramsundarDifferentiablePhysicsPosition2021,
@@ -5798,15 +6429,14 @@
number = {arXiv:2109.07573},
eprint = {2109.07573},
eprinttype = {arxiv},
- primaryclass = {physics},
+ eprintclass = {physics},
publisher = {{arXiv}},
doi = {10.48550/arXiv.2109.07573},
url = {http://arxiv.org/abs/2109.07573},
urldate = {2022-05-18},
abstract = {Differentiable physics provides a new approach for modeling and understanding the physical systems by pairing the new technology of differentiable programming with classical numerical methods for physical simulation. We survey the rapidly growing literature of differentiable physics techniques and highlight methods for parameter estimation, learning representations, solving differential equations, and developing what we call scientific foundation models using data and inductive priors. We argue that differentiable physics offers a new paradigm for modeling physical phenomena by combining classical analytic solutions with numerical methodology using the bridge of differentiable programming.},
- archiveprefix = {arXiv},
keywords = {autodiff,ML,physics-informed ML},
- file = {/home/johannes/Nextcloud/Zotero/Ramsundar et al_2021_Differentiable Physics.pdf;/home/johannes/Zotero/storage/RGUHZPWB/2109.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Ramsundar et al_2021_Differentiable Physics.pdf;/Users/wasmer/Zotero/storage/RGUHZPWB/2109.html}
}
@article{reiserGraphNeuralNetworks2021,
@@ -5824,26 +6454,23 @@
abstract = {Graph neural networks are a versatile machine learning architecture that received a lot of attention recently due to its wide range of applications. In this technical report, we present an implementation of graph convolution and graph pooling layers for TensorFlow-Keras models, which allows a seamless and flexible integration into standard Keras layers to set up graph models in a functional way. We developed the Keras Graph Convolutional Neural Network Python package kgcnn based on TensorFlow-Keras which focus on a transparent tensor structure passed between layers and an ease-of-use mindset.},
langid = {english},
keywords = {GCN,GNN,keras,library,ML,models,SchNet,with-code},
- file = {/home/johannes/Nextcloud/Zotero/Reiser et al_2021_Graph neural networks in TensorFlow-Keras with RaggedTensor representation.pdf;/home/johannes/Zotero/storage/6KMNG399/S266596382100035X.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Reiser et al_2021_Graph neural networks in TensorFlow-Keras with RaggedTensor representation.pdf;/Users/wasmer/Zotero/storage/6KMNG399/S266596382100035X.html}
}
-@misc{reiserGraphNeuralNetworks2022,
+@online{reiserGraphNeuralNetworks2022,
title = {Graph Neural Networks for Materials Science and Chemistry},
- author = {Reiser, Patrick and Neubert, Marlen and Eberhard, André and Torresi, Luca and Zhou, Chen and Shao, Chen and Metni, Houssam and van Hoesel, Clint and Schopmans, Henrik and Sommer, Timo and Friederich, Pascal},
- options = {useprefix=true},
+ author = {Reiser, Patrick and Neubert, Marlen and Eberhard, André and Torresi, Luca and Zhou, Chen and Shao, Chen and Metni, Houssam and family=Hoesel, given=Clint, prefix=van, useprefix=true and Schopmans, Henrik and Sommer, Timo and Friederich, Pascal},
date = {2022-08-05},
number = {arXiv:2208.09481},
- eprint = {2208.09481},
+ eprint = {arXiv:2208.09481},
eprinttype = {arxiv},
- primaryclass = {cond-mat, physics:physics},
- publisher = {{arXiv}},
doi = {10.48550/arXiv.2208.09481},
url = {http://arxiv.org/abs/2208.09481},
urldate = {2022-09-27},
abstract = {Machine learning plays an increasingly important role in many areas of chemistry and materials science, e.g. to predict materials properties, to accelerate simulations, to design new materials, and to predict synthesis routes of new materials. Graph neural networks (GNNs) are one of the fastest growing classes of machine learning models. They are of particular relevance for chemistry and materials science, as they directly work on a graph or structural representation of molecules and materials and therefore have full access to all relevant information required to characterize materials. In this review article, we provide an overview of the basic principles of GNNs, widely used datasets, and state-of-the-art architectures, followed by a discussion of a wide range of recent applications of GNNs in chemistry and materials science, and concluding with a road-map for the further development and application of GNNs.},
- archiveprefix = {arXiv},
+ pubstate = {preprint},
keywords = {_tablet,GCN,GNN,molecules,review,solids},
- file = {/home/johannes/Nextcloud/Zotero/Reiser et al_2022_Graph neural networks for materials science and chemistry.pdf;/home/johannes/Zotero/storage/IVEGXDHZ/2208.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Reiser et al_2022_Graph neural networks for materials science and chemistry.pdf;/Users/wasmer/Zotero/storage/IVEGXDHZ/2208.html}
}
@article{RiseQuantumMaterials2016,
@@ -5864,13 +6491,51 @@
langid = {english},
keywords = {quantum materials,review},
annotation = {Bandiera\_abtest: a Cg\_type: Nature Research Journals Primary\_atype: Editorial Subject\_term: Condensed-matter physics;History;Quantum physics Subject\_term\_id: condensed-matter-physics;history;quantum-physics},
- file = {/home/johannes/Nextcloud/Zotero/2016_The rise of quantum materials.pdf;/home/johannes/Zotero/storage/YG3UAYEY/nphys3668.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/2016_The rise of quantum materials.pdf;/Users/wasmer/Zotero/storage/YG3UAYEY/nphys3668.html}
+}
+
+@article{rodriguezComparativeStudyDifferent2022,
+ title = {A Comparative Study of Different Machine Learning Methods for Dissipative Quantum Dynamics},
+ author = {RodrÃguez, Luis E. Herrera and Ullah, Arif and Espinosa, Kennet J. Rueda and Dral, Pavlo O. and Kananenka, Alexei A.},
+ date = {2022-11},
+ journaltitle = {Machine Learning: Science and Technology},
+ shortjournal = {Mach. Learn.: Sci. Technol.},
+ volume = {3},
+ number = {4},
+ pages = {045016},
+ publisher = {{IOP Publishing}},
+ issn = {2632-2153},
+ doi = {10.1088/2632-2153/ac9a9d},
+ url = {https://dx.doi.org/10.1088/2632-2153/ac9a9d},
+ urldate = {2023-03-02},
+ abstract = {It has been recently shown that supervised machine learning (ML) algorithms can accurately and efficiently predict long-time population dynamics of dissipative quantum systems given only short-time population dynamics. In the present article we benchmarked 22 ML models on their ability to predict long-time dynamics of a two-level quantum system linearly coupled to harmonic bath. The models include uni- and bidirectional recurrent, convolutional, and fully-connected feedforward artificial neural networks (ANNs) and kernel ridge regression (KRR) with linear and most commonly used nonlinear kernels. Our results suggest that KRR with nonlinear kernels can serve as inexpensive yet accurate way to simulate long-time dynamics in cases where the constant length of input trajectories is appropriate. Convolutional gated recurrent unit model is found to be the most efficient ANN model.},
+ langid = {english},
+ keywords = {/unread,AML,benchmarking,CNN,dynamics,KRR,ML,MLatom,NN,RNN},
+ file = {/Users/wasmer/Nextcloud/Zotero/RodrÃguez et al_2022_A comparative study of different machine learning methods for dissipative.pdf}
+}
+
+@article{ronneAtomisticStructureSearch2022,
+ title = {Atomistic Structure Search Using Local Surrogate Model},
+ author = {Rønne, Nikolaj and Christiansen, Mads-Peter V. and Slavensky, Andreas Møller and Tang, Zeyuan and Brix, Florian and Pedersen, Mikkel Elkjær and Bisbo, Malthe Kjær and Hammer, Bjørk},
+ date = {2022-11-07},
+ journaltitle = {The Journal of Chemical Physics},
+ shortjournal = {J. Chem. Phys.},
+ volume = {157},
+ number = {17},
+ pages = {174115},
+ publisher = {{American Institute of Physics}},
+ issn = {0021-9606},
+ doi = {10.1063/5.0121748},
+ url = {https://aip.scitation.org/doi/full/10.1063/5.0121748},
+ urldate = {2023-01-29},
+ abstract = {We describe a local surrogate model for use in conjunction with global structure search methods. The model follows the Gaussian approximation potential formalism and is based on the smooth overlap of atomic positions descriptor with sparsification in terms of a reduced number of local environments using mini-batch k-means. The model is implemented in the Atomistic Global Optimization X framework and used as a partial replacement of the local relaxations in basin hopping structure search. The approach is shown to be robust for a wide range of atomistic systems, including molecules, nanoparticles, surface supported clusters, and surface thin films. The benefits in a structure search context of a local surrogate model are demonstrated. This includes the ability to benefit from transfer learning from smaller systems as well as the possibility to perform concurrent multi-stoichiometry searches.},
+ keywords = {/unread},
+ file = {/Users/wasmer/Nextcloud/Zotero/Rønne et al_2022_Atomistic structure search using local surrogate model.pdf}
}
@article{ruppFastAccurateModeling2012,
title = {Fast and {{Accurate Modeling}} of {{Molecular Atomization Energies}} with {{Machine Learning}}},
- author = {Rupp, Matthias and Tkatchenko, Alexandre and Müller, Klaus-Robert and von Lilienfeld, O. Anatole},
- options = {useprefix=true},
+ author = {Rupp, Matthias and Tkatchenko, Alexandre and Müller, Klaus-Robert and family=Lilienfeld, given=O. Anatole, prefix=von, useprefix=true},
date = {2012-01-31},
journaltitle = {Physical Review Letters},
shortjournal = {Phys. Rev. Lett.},
@@ -5883,7 +6548,7 @@
urldate = {2021-07-10},
abstract = {We introduce a machine learning model to predict atomization energies of a diverse set of organic molecules, based on nuclear charges and atomic positions only. The problem of solving the molecular Schrödinger equation is mapped onto a nonlinear statistical regression problem of reduced complexity. Regression models are trained on and compared to atomization energies computed with hybrid density-functional theory. Cross validation over more than seven thousand organic molecules yields a mean absolute error of ∼10 kcal/mol. Applicability is demonstrated for the prediction of molecular atomization potential energy curves.},
keywords = {Coulomb matrix,descriptors,ML,original publication},
- file = {/home/johannes/Nextcloud/Zotero/Rupp et al_2012_Fast and Accurate Modeling of Molecular Atomization Energies with Machine.pdf;/home/johannes/Zotero/storage/AP7Y6JEW/PhysRevLett.108.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Rupp et al_2012_Fast and Accurate Modeling of Molecular Atomization Energies with Machine.pdf;/Users/wasmer/Zotero/storage/AP7Y6JEW/PhysRevLett.108.html}
}
@article{ruppMachineLearningQuantum2015,
@@ -5901,8 +6566,7 @@
abstract = {Models that combine quantum mechanics (QM) with machine learning (ML) promise to deliver the accuracy of QM at the speed of ML. This hands-on tutorial introduces the reader to QM/ML models based on kernel learning, an elegant, systematically nonlinear form of ML. Pseudocode and a reference implementation are provided, enabling the reader to reproduce results from recent publications where atomization energies of small organic molecules are predicted using kernel ridge regression. © 2015 Wiley Periodicals, Inc.},
langid = {english},
keywords = {_tablet,Coulomb matrix,GPR,kernel methods,KRR,ML,models,tutorial},
- annotation = {\_eprint: https://onlinelibrary.wiley.com/doi/pdf/10.1002/qua.24954},
- file = {/home/johannes/Nextcloud/Zotero/Rupp_2015_Machine learning for quantum mechanics in a nutshell.pdf;/home/johannes/Zotero/storage/7CP5YBAD/qua.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Rupp_2015_Machine learning for quantum mechanics in a nutshell.pdf;/Users/wasmer/Zotero/storage/7CP5YBAD/qua.html}
}
@article{russmannAiiDAKKRPluginIts2021,
@@ -5922,7 +6586,7 @@
issue = {1},
langid = {english},
keywords = {_tablet,AiiDA,aiida-kkr,defects,FZJ,impurity embedding,juKKR,KKR,PGI-1/IAS-1,physics,topological insulator},
- file = {/home/johannes/Nextcloud/Zotero/Rüßmann et al_2021_The AiiDA-KKR plugin and its application to high-throughput impurity embedding.pdf;/home/johannes/Zotero/storage/X4T36V7Q/s41524-020-00482-5.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Rüßmann et al_2021_The AiiDA-KKR plugin and its application to high-throughput impurity embedding.pdf;/Users/wasmer/Zotero/storage/X4T36V7Q/s41524-020-00482-5.html}
}
@article{russmannAiiDASpiritPluginAutomated2022,
@@ -5937,7 +6601,7 @@
urldate = {2022-08-11},
abstract = {Landau-Lifshitz-Gilbert (LLG) spin-dynamics calculations based on the extended Heisenberg Hamiltonian is an important tool in computational materials science involving magnetic materials. LLG simulations allow to bridge the gap from expensive quantum mechanical calculations with small unit cells to large supercells where the collective behavior of millions of spins can be studied. In this work we present the AiiDA-Spirit plugin that connects the spin-dynamics code Spirit to the AiiDA framework. AiiDA provides a Python interface that facilitates performing high-throughput calculations while automatically augmenting the calculations with metadata describing the data provenance between calculations in a directed acyclic graph. The AiiDA-Spirit interface thus provides an easy way for high-throughput spin-dynamics calculations. The interface to the AiiDA infrastructure furthermore has the advantage that input parameters for the extended Heisenberg model can be extracted from high-throughput first-principles calculations including a proper treatment of the data provenance that ensures reproducibility of the calculation results in accordance to the FAIR principles. We describe the layout of the AiiDA-Spirit plugin and demonstrate its capabilities using selected examples for LLG spin-dynamics and Monte Carlo calculations. Furthermore, the integration with first-principles calculations through AiiDA is demonstrated at the example of γ–Fe, where the complex spin-spiral ground state is investigated.},
keywords = {_tablet,AiiDA,aiida-kkr,Heisenberg model,Jij,KKR,library,PGI-1/IAS-1,rec-by-ruess,spin dynamics,Spirit,with-code},
- file = {/home/johannes/Nextcloud/Zotero/Rüßmann et al_2022_The AiiDA-Spirit Plugin for Automated Spin-Dynamics Simulations and Multi-Scale.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Rüßmann et al_2022_The AiiDA-Spirit Plugin for Automated Spin-Dynamics Simulations and Multi-Scale.pdf}
}
@article{russmannInitioTheoryFourierTransformed2021,
@@ -5955,8 +6619,7 @@
abstract = {The quasiparticle interference (QPI) technique is a powerful tool that allows to uncover the structure and properties of electronic structure of a material combined with scattering properties of defects at surfaces. Recently, this technique has been pivotal in proving the unique properties of the surface state of topological insulators which manifests itself in the absence of backscattering. Herein, a Green function-based formalism is derived for the ab initio computation of Fourier-transformed QPI images. The efficiency of the new implementation is shown at the examples of QPI that forms around magnetic and nonmagnetic defects at the Bi2Te3 surface. This method allows a deepened understanding of the scattering properties of topologically protected electrons off defects and is a useful tool in the study of quantum materials in the future.},
langid = {english},
keywords = {_tablet,density functional theory,impurity scattering,Korringa–Kohn–Rostoker,quasiparticle interferences,topological insulators},
- annotation = {\_eprint: https://onlinelibrary.wiley.com/doi/pdf/10.1002/pssb.202000031},
- file = {/home/johannes/Nextcloud/Zotero/Rüßmann et al_2021_Ab Initio Theory of Fourier-Transformed Quasiparticle Interference Maps and.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Rüßmann et al_2021_Ab Initio Theory of Fourier-Transformed Quasiparticle Interference Maps and.pdf}
}
@thesis{russmannSpinScatteringTopologically2018,
@@ -5971,7 +6634,7 @@
isbn = {9783958063365},
langid = {english},
keywords = {juKKR,KKR,PGI-1/IAS-1,thesis},
- file = {/home/johannes/Nextcloud/Zotero/false;/home/johannes/Zotero/storage/T7V45S9S/850306.html}
+ file = {/home/johannes/Nextcloud/Zotero/false;/Users/wasmer/Zotero/storage/T7V45S9S/850306.html}
}
@article{ryczkoDeepLearningDensityfunctional2019,
@@ -5984,7 +6647,7 @@
number = {2},
doi = {10.1103/PhysRevA.100.022512},
keywords = {CNN,Condensed Matter - Materials Science,DFT,featureless,ML,ML-DFT,ML-ESM,models,Physics - Computational Physics,rec-by-bluegel},
- file = {/home/johannes/Nextcloud/Zotero/Ryczko_2019_Deep learning and density-functional theory.pdf;/home/johannes/Zotero/storage/DYNSZ4CL/1811.html;/home/johannes/Zotero/storage/NJ67I8R7/PhysRevA.100.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Ryczko_2019_Deep learning and density-functional theory.pdf;/Users/wasmer/Zotero/storage/DYNSZ4CL/1811.html;/Users/wasmer/Zotero/storage/NJ67I8R7/PhysRevA.100.html}
}
@article{saalMaterialsDesignDiscovery2013,
@@ -6003,7 +6666,7 @@
urldate = {2021-10-15},
abstract = {High-throughput density functional theory (HT DFT) is fast becoming a powerful tool for accelerating materials design and discovery by the amassing tens and even hundreds of thousands of DFT calculations in large databases. Complex materials problems can be approached much more efficiently and broadly through the sheer quantity of structures and chemistries available in such databases. Our HT DFT database, the Open Quantum Materials Database (OQMD), contains over 200,000 DFT calculated crystal structures and will be freely available for public use at http://oqmd.org. In this review, we describe the OQMD and its use in five materials problems, spanning a wide range of applications and materials types: (I) Li-air battery combination catalyst/electrodes, (II) Li-ion battery anodes, (III) Li-ion battery cathode coatings reactive with HF, (IV) Mg-alloy long-period stacking ordered (LPSO) strengthening precipitates, and (V) training a machine learning model to predict new stable ternary compounds.},
langid = {english},
- file = {/home/johannes/Nextcloud/Zotero/Saal et al_2013_Materials Design and Discovery with High-Throughput Density Functional Theory.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Saal et al_2013_Materials Design and Discovery with High-Throughput Density Functional Theory.pdf}
}
@unpublished{samuelMachineLearningPipelines2020,
@@ -6013,13 +6676,12 @@
date = {2020-06-22},
eprint = {2006.12117},
eprinttype = {arxiv},
- primaryclass = {cs, stat},
+ eprintclass = {cs, stat},
url = {http://arxiv.org/abs/2006.12117},
urldate = {2021-10-21},
abstract = {Machine learning (ML) is an increasingly important scientific tool supporting decision making and knowledge generation in numerous fields. With this, it also becomes more and more important that the results of ML experiments are reproducible. Unfortunately, that often is not the case. Rather, ML, similar to many other disciplines, faces a reproducibility crisis. In this paper, we describe our goals and initial steps in supporting the end-to-end reproducibility of ML pipelines. We investigate which factors beyond the availability of source code and datasets influence reproducibility of ML experiments. We propose ways to apply FAIR data practices to ML workflows. We present our preliminary results on the role of our tool, ProvBook, in capturing and comparing provenance of ML experiments and their reproducibility using Jupyter Notebooks.},
- archiveprefix = {arXiv},
keywords = {Computer Science - Artificial Intelligence,Computer Science - Machine Learning,Statistics - Machine Learning},
- file = {/home/johannes/Nextcloud/Zotero/Samuel et al_2020_Machine Learning Pipelines.pdf;/home/johannes/Zotero/storage/GTJD4NAB/2006.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Samuel et al_2020_Machine Learning Pipelines.pdf;/Users/wasmer/Zotero/storage/GTJD4NAB/2006.html}
}
@inproceedings{satorrasEquivariantGraphNeural2021,
@@ -6036,7 +6698,7 @@
eventtitle = {International {{Conference}} on {{Machine Learning}}},
langid = {english},
keywords = {EGNN,equivariant,GDL,GNN,ML,NN,original publication,rec-by-bluegel},
- file = {/home/johannes/Nextcloud/Zotero/Satorras et al_2021_E(n) Equivariant Graph Neural Networks2.pdf;/home/johannes/Zotero/storage/3ATM3ZJA/Satorras et al_2021_E(n) Equivariant Graph Neural Networks.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Satorras et al_2021_E(n) Equivariant Graph Neural Networks2.pdf;/Users/wasmer/Zotero/storage/3ATM3ZJA/Satorras et al_2021_E(n) Equivariant Graph Neural Networks.pdf}
}
@article{saucedaBIGDMLAccurateQuantum2022,
@@ -6057,7 +6719,7 @@
issue = {1},
langid = {english},
keywords = {/unread,Coulomb matrix,crystal symmetry,defects,force fields,interfaces and thin films,library,materials,MD,ML,MLFF,MLP,models,molecular dynamics,PyTorch,sGDML,surface physics,symmetry,with-code},
- file = {/home/johannes/Nextcloud/Zotero/Sauceda et al_2022_BIGDML—Towards accurate quantum machine learning force fields for materials.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Sauceda et al_2022_BIGDML—Towards accurate quantum machine learning force fields for materials.pdf}
}
@unpublished{saucedaBIGDMLExactMachine2021,
@@ -6067,13 +6729,12 @@
date = {2021-06-08},
eprint = {2106.04229},
eprinttype = {arxiv},
- primaryclass = {cond-mat, physics:physics, physics:quant-ph},
+ eprintclass = {cond-mat, physics:physics, physics:quant-ph},
url = {http://arxiv.org/abs/2106.04229},
urldate = {2021-06-17},
abstract = {Machine-learning force fields (MLFF) should be accurate, computationally and data efficient, and applicable to molecules, materials, and interfaces thereof. Currently, MLFFs often introduce tradeoffs that restrict their practical applicability to small subsets of chemical space or require exhaustive datasets for training. Here, we introduce the Bravais-Inspired Gradient-Domain Machine Learning (BIGDML) approach and demonstrate its ability to construct reliable force fields using a training set with just 10-200 geometries for materials including pristine and defect-containing 2D and 3D semiconductors and metals, as well as chemisorbed and physisorbed atomic and molecular adsorbates on surfaces. The BIGDML model employs the full relevant symmetry group for a given material, does not assume artificial atom types or localization of atomic interactions and exhibits high data efficiency and state-of-the-art energy accuracies (errors substantially below 1 meV per atom) for an extended set of materials. Extensive path-integral molecular dynamics carried out with BIGDML models demonstrate the counterintuitive localization of benzene--graphene dynamics induced by nuclear quantum effects and allow to rationalize the Arrhenius behavior of hydrogen diffusion coefficient in a Pd crystal for a wide range of temperatures.},
- archiveprefix = {arXiv},
keywords = {CM,ML,ML-FF,MLP,models,PES,sGDML},
- file = {/home/johannes/Nextcloud/Zotero/Sauceda et al_2021_BIGDML.pdf;/home/johannes/Zotero/storage/XVR5SBVI/2106.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Sauceda et al_2021_BIGDML.pdf;/Users/wasmer/Zotero/storage/XVR5SBVI/2106.html}
}
@unpublished{scherbelaSolvingElectronicSchr2021,
@@ -6082,13 +6743,12 @@
date = {2021-12-17},
eprint = {2105.08351},
eprinttype = {arxiv},
- primaryclass = {physics},
+ eprintclass = {physics},
url = {http://arxiv.org/abs/2105.08351},
urldate = {2022-03-28},
abstract = {Accurate numerical solutions for the Schr\textbackslash "odinger equation are of utmost importance in quantum chemistry. However, the computational cost of current high-accuracy methods scales poorly with the number of interacting particles. Combining Monte Carlo methods with unsupervised training of neural networks has recently been proposed as a promising approach to overcome the curse of dimensionality in this setting and to obtain accurate wavefunctions for individual molecules at a moderately scaling computational cost. These methods currently do not exploit the regularity exhibited by wavefunctions with respect to their molecular geometries. Inspired by recent successful applications of deep transfer learning in machine translation and computer vision tasks, we attempt to leverage this regularity by introducing a weight-sharing constraint when optimizing neural network-based models for different molecular geometries. That is, we restrict the optimization process such that up to 95 percent of weights in a neural network model are in fact equal across varying molecular geometries. We find that this technique can accelerate optimization when considering sets of nuclear geometries of the same molecule by an order of magnitude and that it opens a promising route towards pre-trained neural network wavefunctions that yield high accuracy even across different molecules.},
- archiveprefix = {arXiv},
keywords = {Computer Science - Machine Learning,Physics - Chemical Physics,Physics - Computational Physics},
- file = {/home/johannes/Nextcloud/Zotero/Scherbela et al_2021_Solving the electronic Schr-odinger equation for multiple nuclear geometries.pdf;/home/johannes/Zotero/storage/PSDVAEAB/2105.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Scherbela et al_2021_Solving the electronic Schr-odinger equation for multiple nuclear geometries.pdf;/Users/wasmer/Zotero/storage/PSDVAEAB/2105.html}
}
@article{scherbelaSolvingElectronicSchrodinger2022,
@@ -6109,7 +6769,27 @@
issue = {5},
langid = {english},
keywords = {Backflow,cusps,DeepErwin,FermiNet,JAX,library,MCMC,ML-ESM,ML-QMBP,molecules,PauliNet,prediction of wavefunction,QMC,VMC,weight-sharing,with-code},
- file = {/home/johannes/Nextcloud/Zotero/Scherbela et al_2022_Solving the electronic Schrödinger equation for multiple nuclear geometries.pdf;/home/johannes/Zotero/storage/JHT752NC/s43588-022-00228-x.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Scherbela et al_2022_Solving the electronic Schrödinger equation for multiple nuclear geometries.pdf;/Users/wasmer/Zotero/storage/JHT752NC/s43588-022-00228-x.html}
+}
+
+@article{schlenzApplicationsMachineLearning2022,
+ title = {Applications of {{Machine Learning}} to the {{Study}} of {{Crystalline Materials}}},
+ author = {Schlenz, Hartmut and Sandfeld, Stefan},
+ date = {2022-08},
+ journaltitle = {Crystals},
+ volume = {12},
+ number = {8},
+ pages = {1070},
+ publisher = {{Multidisciplinary Digital Publishing Institute}},
+ issn = {2073-4352},
+ doi = {10.3390/cryst12081070},
+ url = {https://www.mdpi.com/2073-4352/12/8/1070},
+ urldate = {2023-03-02},
+ abstract = {This Special Issue, “Applications of Machine Learning to the Study of Crystalline Materialsâ€, is a collection of seven original articles published in 2021 and 2022 and dedicated to applications of machine learning in materials research [...]},
+ issue = {8},
+ langid = {english},
+ keywords = {/unread,FZJ,IEK-1,materials,n/a},
+ file = {/Users/wasmer/Nextcloud/Zotero/Schlenz_Sandfeld_2022_Applications of Machine Learning to the Study of Crystalline Materials.pdf}
}
@article{schmidtMachineLearningPhysical2019,
@@ -6127,7 +6807,7 @@
urldate = {2022-07-05},
abstract = {We train a neural network as the universal exchange–correlation functional of density-functional theory that simultaneously reproduces both the exact exchange–correlation energy and the potential. This functional is extremely nonlocal but retains the computational scaling of traditional local or semilocal approximations. It therefore holds the promise of solving some of the delocalization problems that plague density-functional theory, while maintaining the computational efficiency that characterizes the Kohn–Sham equations. Furthermore, by using automatic differentiation, a capability present in modern machine-learning frameworks, we impose the exact mathematical relation between the exchange–correlation energy and the potential, leading to a fully consistent method. We demonstrate the feasibility of our approach by looking at one-dimensional systems with two strongly correlated electrons, where density-functional methods are known to fail, and investigate the behavior and performance of our functional by varying the degree of nonlocality.},
keywords = {autodiff,DFT,ML,ML-DFA,ML-DFT,ML-ESM,prediction from density,prediction of Exc,prediction of vxc,pytorch},
- file = {/home/johannes/Nextcloud/Zotero/Schmidt et al_2019_Machine Learning the Physical Nonlocal Exchange–Correlation Functional of.pdf;/home/johannes/Zotero/storage/QCMK7FSR/acs.jpclett.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Schmidt et al_2019_Machine Learning the Physical Nonlocal Exchange–Correlation Functional of.pdf;/Users/wasmer/Zotero/storage/QCMK7FSR/acs.jpclett.html}
}
@article{schmidtRecentAdvancesApplications2019,
@@ -6148,13 +6828,12 @@
issue = {1},
langid = {english},
annotation = {Bandiera\_abtest: a Cc\_license\_type: cc\_by Cg\_type: Nature Research Journals Primary\_atype: Reviews Subject\_term: Condensed-matter physics;Electronic structure;Materials science;Metals and alloys;Semiconductors Subject\_term\_id: condensed-matter-physics;electronic-structure;materials-science;metals-and-alloys;semiconductors},
- file = {/home/johannes/Nextcloud/Zotero/Schmidt et al_2019_Recent advances and applications of machine learning in solid-state materials.pdf;/home/johannes/Zotero/storage/BY9RESIZ/s41524-019-0221-0.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Schmidt et al_2019_Recent advances and applications of machine learning in solid-state materials.pdf;/Users/wasmer/Zotero/storage/BY9RESIZ/s41524-019-0221-0.html}
}
@article{schmittIntegrationTopologicalInsulator2022,
title = {Integration of {{Topological Insulator Josephson Junctions}} in {{Superconducting Qubit Circuits}}},
- author = {Schmitt, Tobias W. and Connolly, Malcolm R. and Schleenvoigt, Michael and Liu, Chenlu and Kennedy, Oscar and Chávez-Garcia, José M. and Jalil, Abdur R. and Bennemann, Benjamin and Trellenkamp, Stefan and Lentz, Florian and Neumann, Elmar and Lindström, Tobias and de Graaf, Sebastian E. and Berenschot, Erwin and Tas, Niels and Mussler, Gregor and Petersson, Karl D. and Grützmacher, Detlev and Schüffelgen, Peter},
- options = {useprefix=true},
+ author = {Schmitt, Tobias W. and Connolly, Malcolm R. and Schleenvoigt, Michael and Liu, Chenlu and Kennedy, Oscar and Chávez-Garcia, José M. and Jalil, Abdur R. and Bennemann, Benjamin and Trellenkamp, Stefan and Lentz, Florian and Neumann, Elmar and Lindström, Tobias and family=Graaf, given=Sebastian E., prefix=de, useprefix=true and Berenschot, Erwin and Tas, Niels and Mussler, Gregor and Petersson, Karl D. and Grützmacher, Detlev and Schüffelgen, Peter},
date = {2022-04-13},
journaltitle = {Nano Letters},
shortjournal = {Nano Lett.},
@@ -6168,7 +6847,7 @@
urldate = {2022-04-16},
abstract = {The integration of semiconductor Josephson junctions (JJs) in superconducting quantum circuits provides a versatile platform for hybrid qubits and offers a powerful way to probe exotic quasiparticle excitations. Recent proposals for using circuit quantum electrodynamics (cQED) to detect topological superconductivity motivate the integration of novel topological materials in such circuits. Here, we report on the realization of superconducting transmon qubits implemented with (Bi0.06Sb0.94)2Te3 topological insulator (TI) JJs using ultrahigh vacuum fabrication techniques. Microwave losses on our substrates, which host monolithically integrated hardmasks used for the selective area growth of TI nanostructures, imply microsecond limits to relaxation times and, thus, their compatibility with strong-coupling cQED. We use the cavity–qubit interaction to show that the Josephson energy of TI-based transmons scales with their JJ dimensions and demonstrate qubit control as well as temporal quantum coherence. Our results pave the way for advanced investigations of topological materials in both novel Josephson and topological qubits.},
keywords = {_tablet,experimental,FZJ,PGI,PGI-9,quantum computing,superconductor,topological insulator},
- file = {/home/johannes/Nextcloud/Zotero/Schmitt et al_2022_Integration of Topological Insulator Josephson Junctions in Superconducting.pdf;/home/johannes/Zotero/storage/ZVNVRDSF/acs.nanolett.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Schmitt et al_2022_Integration of Topological Insulator Josephson Junctions in Superconducting.pdf;/Users/wasmer/Zotero/storage/ZVNVRDSF/acs.nanolett.html}
}
@article{schuchComputationalComplexityInteracting2009,
@@ -6189,13 +6868,12 @@
issue = {10},
langid = {english},
keywords = {computational complexity,DFT,NP-hard,Schrödinger equation},
- file = {/home/johannes/Nextcloud/Zotero/Schuch_Verstraete_2009_Computational complexity of interacting electrons and fundamental limitations.pdf;/home/johannes/Zotero/storage/3TCEDPTF/nphys1370.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Schuch_Verstraete_2009_Computational complexity of interacting electrons and fundamental limitations.pdf;/Users/wasmer/Zotero/storage/3TCEDPTF/nphys1370.html}
}
@book{schuttMachineLearningMeets2020,
title = {Machine {{Learning Meets Quantum Physics}}},
- editor = {Schütt, Kristof T. and Chmiela, Stefan and von Lilienfeld, O. Anatole and Tkatchenko, Alexandre and Tsuda, Koji and Müller, Klaus-Robert},
- options = {useprefix=true},
+ editor = {Schütt, Kristof T. and Chmiela, Stefan and family=Lilienfeld, given=O. Anatole, prefix=von, useprefix=true and Tkatchenko, Alexandre and Tsuda, Koji and Müller, Klaus-Robert},
date = {2020},
series = {Lecture {{Notes}} in {{Physics}}},
volume = {968},
@@ -6207,7 +6885,7 @@
isbn = {978-3-030-40244-0 978-3-030-40245-7},
langid = {english},
keywords = {chemistry,descriptors,general,ML,models,physics,review},
- file = {/home/johannes/Books/scientific_machine_learning/Schütt et al_2020_Machine Learning Meets Quantum Physics.pdf}
+ file = {/Users/wasmer/Books/scientific_machine_learning/Schütt et al_2020_Machine Learning Meets Quantum Physics.pdf}
}
@article{schuttSchNetDeepLearning2018,
@@ -6226,26 +6904,24 @@
urldate = {2022-10-03},
abstract = {Deep learning has led to a paradigm shift in artificial intelligence, including web, text, and image search, speech recognition, as well as bioinformatics, with growing impact in chemical physics. Machine learning, in general, and deep learning, in particular, are ideally suitable for representing quantum-mechanical interactions, enabling us to model nonlinear potential-energy surfaces or enhancing the exploration of chemical compound space. Here we present the deep learning architecture SchNet that is specifically designed to model atomistic systems by making use of continuous-filter convolutional layers. We demonstrate the capabilities of SchNet by accurately predicting a range of properties across chemical space for molecules and materials, where our model learns chemically plausible embeddings of atom types across the periodic table. Finally, we employ SchNet to predict potential-energy surfaces and energy-conserving force fields for molecular dynamics simulations of small molecules and perform an exemplary study on the quantum-mechanical properties of C20-fullerene that would have been infeasible with regular ab initio molecular dynamics.},
keywords = {DTNN,MD,ML,MLP,NN,SchNet},
- file = {/home/johannes/Nextcloud/Zotero/Schütt et al_2018_SchNet – A deep learning architecture for molecules and materials.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Schütt et al_2018_SchNet – A deep learning architecture for molecules and materials.pdf}
}
-@misc{schuttSchNetPackNeuralNetwork2022,
+@online{schuttSchNetPackNeuralNetwork2022,
title = {{{SchNetPack}} 2.0: {{A}} Neural Network Toolbox for Atomistic Machine Learning},
shorttitle = {{{SchNetPack}} 2.0},
author = {Schütt, Kristof T. and Hessmann, Stefaan S. P. and Gebauer, Niklas W. A. and Lederer, Jonas and Gastegger, Michael},
date = {2022-12-11},
number = {arXiv:2212.05517},
- eprint = {2212.05517},
+ eprint = {arXiv:2212.05517},
eprinttype = {arxiv},
- primaryclass = {physics, stat},
- publisher = {{arXiv}},
doi = {10.48550/arXiv.2212.05517},
url = {http://arxiv.org/abs/2212.05517},
urldate = {2022-12-27},
abstract = {SchNetPack is a versatile neural networks toolbox that addresses both the requirements of method development and application of atomistic machine learning. Version 2.0 comes with an improved data pipeline, modules for equivariant neural networks as well as a PyTorch implementation of molecular dynamics. An optional integration with PyTorch Lightning and the Hydra configuration framework powers a flexible command-line interface. This makes SchNetPack 2.0 easily extendable with custom code and ready for complex training task such as generation of 3d molecular structures.},
- archiveprefix = {arXiv},
- keywords = {/unread,Deep learning,equivariant,Hydra,library,MLP,models,PAiNN,pytorch,SchNet,SO(3),with-code},
- file = {/home/johannes/Nextcloud/Zotero/Schütt et al_2022_SchNetPack 2.pdf;/home/johannes/Zotero/storage/AHBKQSBM/2212.html}
+ pubstate = {preprint},
+ keywords = {_tablet,Deep learning,equivariant,Hydra,library,MLP,models,PAiNN,pytorch,SchNet,SO(3),with-code},
+ file = {/Users/wasmer/Nextcloud/Zotero/Schütt et al_2022_SchNetPack 2.pdf;/Users/wasmer/Zotero/storage/AHBKQSBM/2212.html}
}
@article{schuttUnifyingMachineLearning2019,
@@ -6266,7 +6942,7 @@
issue = {1},
langid = {english},
keywords = {ANN,ML,ML-ESM,models,molecules,original publication,prediction of wavefunction,SchNet,SchNOrb,WFT},
- file = {/home/johannes/Nextcloud/Zotero/Schütt et al_2019_Unifying machine learning and quantum chemistry with a deep neural network for.pdf;/home/johannes/Zotero/storage/ADRZDHRZ/s41467-019-12875-2.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Schütt et al_2019_Unifying machine learning and quantum chemistry with a deep neural network for.pdf;/Users/wasmer/Zotero/storage/ADRZDHRZ/s41467-019-12875-2.html}
}
@article{sendekMachineLearningModeling,
@@ -6283,26 +6959,39 @@
abstract = {Machine learning (ML)-based approaches to battery design are relatively new but demonstrate significant promise for accelerating the timeline for new materials discovery, process optimization, and cell lifetime prediction. Battery modeling represents an interesting and unconventional application area for ML, as datasets are often small but some degree of physical understanding of the underlying processes may exist. This review article provides discussion and analysis of several important and increasingly common questions: how ML-based battery modeling works, how much data are required, how to judge model performance, and recommendations for building models in the small data regime. This article begins with an introduction to ML in general, highlighting several important concepts for small data applications. Previous ionic conductivity modeling efforts are discussed in depth as a case study to illustrate these modeling concepts. Finally, an overview of modeling efforts in major areas of battery design is provided and several areas for promising future efforts are identified, within the context of typical small data constraints.},
langid = {english},
keywords = {_tablet,chemistry,materials informatics,ML,small data,tutorial},
- annotation = {\_eprint: https://onlinelibrary.wiley.com/doi/pdf/10.1002/aenm.202200553},
- file = {/home/johannes/Nextcloud/Zotero/Sendek et al_Machine Learning Modeling for Accelerated Battery Materials Design in the Small.pdf;/home/johannes/Zotero/storage/55KE647F/aenm.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Sendek et al_Machine Learning Modeling for Accelerated Battery Materials Design in the Small.pdf;/Users/wasmer/Zotero/storage/55KE647F/aenm.html}
+}
+
+@online{shahPhysicsInformedNeuralNetworks2022,
+ title = {Physics-{{Informed Neural Networks}} as {{Solvers}} for the {{Time-Dependent Schr}}\textbackslash "odinger {{Equation}}},
+ author = {Shah, Karan and Stiller, Patrick and Hoffmann, Nico and Cangi, Attila},
+ date = {2022-10-22},
+ number = {arXiv:2210.12522},
+ eprint = {arXiv:2210.12522},
+ eprinttype = {arxiv},
+ doi = {10.48550/arXiv.2210.12522},
+ url = {http://arxiv.org/abs/2210.12522},
+ urldate = {2023-02-15},
+ abstract = {We demonstrate the utility of physics-informed neural networks (PINNs) as solvers for the non-relativistic, time-dependent Schr\textbackslash "odinger equation. We study the performance and generalisability of PINN solvers on the time evolution of a quantum harmonic oscillator across varying system parameters, domains, and energy states.},
+ pubstate = {preprint},
+ keywords = {/unread},
+ file = {/Users/wasmer/Nextcloud/Zotero/Shah et al_2022_Physics-Informed Neural Networks as Solvers for the Time-Dependent.pdf;/Users/wasmer/Zotero/storage/NSJSIKTH/2210.html}
}
-@misc{shenRepresentationindependentElectronicCharge2021,
+@online{shenRepresentationindependentElectronicCharge2021,
title = {A Representation-Independent Electronic Charge Density Database for Crystalline Materials},
author = {Shen, Jimmy-Xuan and Munro, Jason M. and Horton, Matthew K. and Huck, Patrick and Dwaraknath, Shyam and Persson, Kristin A.},
date = {2021-07-07},
number = {arXiv:2107.03540},
- eprint = {2107.03540},
+ eprint = {arXiv:2107.03540},
eprinttype = {arxiv},
- primaryclass = {cond-mat},
- publisher = {{arXiv}},
doi = {10.48550/arXiv.2107.03540},
url = {http://arxiv.org/abs/2107.03540},
urldate = {2022-12-31},
abstract = {In addition to being the core quantity in density functional theory, the charge density can be used in many tertiary analyses in materials sciences from bonding to assigning charge to specific atoms. The charge density is data-rich since it contains information about all the electrons in the system. With increasing utilization of machine-learning tools in materials sciences, a data-rich object like the charge density can be utilized in a wide range of applications. The database presented here provides a modern and user-friendly interface for a large and continuously updated collection of charge densities as part of the Materials Project. In addition to the charge density data, we provide the theory and code for changing the representation of the charge density which should enable more advanced machine-learning studies for the broader community.},
- archiveprefix = {arXiv},
- keywords = {/unread,charge density,data repositories,Database,dimensionality reduction of target,electronic structure,library,materials,materials database,materials project,ML,ML-DFT,prediction from density,prediction of electron density,with-code},
- file = {/home/johannes/Nextcloud/Zotero/Shen et al_2021_A representation-independent electronic charge density database for crystalline.pdf;/home/johannes/Zotero/storage/9A3MUVVK/2107.html}
+ pubstate = {preprint},
+ keywords = {_tablet,/unread,charge density,data repositories,Database,dimensionality reduction of target,electronic structure,library,materials,materials database,materials project,ML,ML-DFT,prediction from density,prediction of electron density,with-code},
+ file = {/Users/wasmer/Nextcloud/Zotero/Shen et al_2021_A representation-independent electronic charge density database for crystalline.pdf;/Users/wasmer/Zotero/storage/9A3MUVVK/2107.html}
}
@article{shmilovichOrbitalMixerUsing2022,
@@ -6318,8 +7007,8 @@
url = {https://doi.org/10.1021/acs.jctc.2c00555},
urldate = {2022-09-29},
abstract = {Leveraging ab initio data at scale has enabled the development of machine learning models capable of extremely accurate and fast molecular property prediction. A central paradigm of many previous studies focuses on generating predictions for only a fixed set of properties. Recent lines of research instead aim to explicitly learn the electronic structure via molecular wavefunctions, from which other quantum chemical properties can be directly derived. While previous methods generate predictions as a function of only the atomic configuration, in this work we present an alternate approach that directly purposes basis-dependent information to predict molecular electronic structure. Our model, Orbital Mixer, is composed entirely of multi-layer perceptrons (MLPs) using MLP-Mixer layers within a simple, intuitive, and scalable architecture that achieves competitive Hamiltonian and molecular orbital energy and coefficient prediction accuracies compared to the state-of-the-art.},
- keywords = {ML,ML-ESM,MLP,molecules,Orbital Mixer,original publication,PhiSNet,prediction of wavefunction,SchNOrb,with-code},
- file = {/home/johannes/Nextcloud/Zotero/Shmilovich et al_2022_Orbital Mixer.pdf}
+ keywords = {_tablet,ML,ML-ESM,MLP,molecules,Orbital Mixer,original publication,PhiSNet,prediction of wavefunction,SchNOrb,with-code},
+ file = {/Users/wasmer/Nextcloud/Zotero/Shmilovich et al_2022_Orbital Mixer.pdf}
}
@article{simmhanSurveyDataProvenance2005,
@@ -6336,7 +7025,7 @@
url = {https://doi.org/10.1145/1084805.1084812},
urldate = {2021-10-17},
abstract = {Data management is growing in complexity as large-scale applications take advantage of the loosely coupled resources brought together by grid middleware and by abundant storage capacity. Metadata describing the data products used in and generated by these applications is essential to disambiguate the data and enable reuse. Data provenance, one kind of metadata, pertains to the derivation history of a data product starting from its original sources.In this paper we create a taxonomy of data provenance characteristics and apply it to current research efforts in e-science, focusing primarily on scientific workflow approaches. The main aspect of our taxonomy categorizes provenance systems based on why they record provenance, what they describe, how they represent and store provenance, and ways to disseminate it. The survey culminates with an identification of open research problems in the field.},
- file = {/home/johannes/Nextcloud/Zotero/Simmhan et al_2005_A survey of data provenance in e-science.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Simmhan et al_2005_A survey of data provenance in e-science.pdf}
}
@article{singraberParallelMultistreamTraining2019,
@@ -6355,7 +7044,7 @@
urldate = {2021-05-18},
abstract = {Over the past years high-dimensional neural network potentials (HDNNPs), fitted to accurately reproduce ab initio potential energy surfaces, have become a powerful tool in chemistry, physics and materials science. Here, we focus on the training of the neural networks that lies at the heart of the HDNNP method. We present an efficient approach for optimizing the weight parameters of the neural network via multistream Kalman filtering, using potential energies and forces as reference data. In this procedure, the choice of the free parameters of the Kalman filter can have a significant impact on the fit quality. Carrying out a large parameter study, we determine optimal settings and demonstrate how to optimize training results of HDNNPs. Moreover, we illustrate our HDNNP training approach by revisiting previously presented fits for water and developing a new potential for copper sulfide. This material, accessible in computer simulations so far only via first-principles methods, forms a particularly complex solid structure at low temperatures and undergoes a phase transition to a superionic state upon heating. Analyzing MD simulations carried out with the Cu2S HDNNP, we confirm that the underlying ab initio reference method indeed reproduces this behavior.},
keywords = {HDNNP,ML,MLP,models,parallelization},
- file = {/home/johannes/Nextcloud/Zotero/Singraber et al_2019_Parallel Multistream Training of High-Dimensional Neural Network Potentials.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Singraber et al_2019_Parallel Multistream Training of High-Dimensional Neural Network Potentials.pdf}
}
@article{sivaramanMachinelearnedInteratomicPotentials2020,
@@ -6376,7 +7065,7 @@
issue = {1},
langid = {english},
keywords = {active learning,AIMD,GAP,MD,ML},
- file = {/home/johannes/Nextcloud/Zotero/Sivaraman et al_2020_Machine-learned interatomic potentials by active learning.pdf;/home/johannes/Zotero/storage/WMMU2G78/s41524-020-00367-7.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Sivaraman et al_2020_Machine-learned interatomic potentials by active learning.pdf;/Users/wasmer/Zotero/storage/WMMU2G78/s41524-020-00367-7.html}
}
@article{smidtFindingSymmetryBreaking2021,
@@ -6394,7 +7083,7 @@
urldate = {2022-10-17},
abstract = {Curie's principle states that “when effects show certain asymmetry, this asymmetry must be found in the causes that gave rise to them.†We demonstrate that symmetry equivariant neural networks uphold Curie's principle and can be used to articulate many symmetry-relevant scientific questions as simple optimization problems. We prove these properties mathematically and demonstrate them numerically by training a Euclidean symmetry equivariant neural network to learn symmetry breaking input to deform a square into a rectangle and to generate octahedra tilting patterns in perovskites.},
keywords = {TODO},
- file = {/home/johannes/Nextcloud/Zotero/Smidt et al_2021_Finding symmetry breaking order parameters with Euclidean neural networks.pdf;/home/johannes/Zotero/storage/TUYBWD9X/Smidt et al. - 2021 - Finding symmetry breaking order parameters with Eu.pdf;/home/johannes/Zotero/storage/4XA7SRBQ/PhysRevResearch.3.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Smidt et al_2021_Finding symmetry breaking order parameters with Euclidean neural networks.pdf;/Users/wasmer/Zotero/storage/TUYBWD9X/Smidt et al. - 2021 - Finding symmetry breaking order parameters with Eu.pdf;/Users/wasmer/Zotero/storage/4XA7SRBQ/PhysRevResearch.3.html}
}
@inproceedings{smithMachineLearningBazaar2020,
@@ -6412,7 +7101,7 @@
eventtitle = {{{SIGMOD}}/{{PODS}} '20: {{International Conference}} on {{Management}} of {{Data}}},
isbn = {978-1-4503-6735-6},
langid = {english},
- file = {/home/johannes/Nextcloud/Zotero/Smith et al_2020_The Machine Learning Bazaar.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Smith et al_2020_The Machine Learning Bazaar.pdf}
}
@article{snyderFindingDensityFunctionals2012,
@@ -6430,7 +7119,7 @@
urldate = {2021-10-15},
abstract = {Machine learning is used to approximate density functionals. For the model problem of the kinetic energy of noninteracting fermions in 1D, mean absolute errors below 1 kcal/mol on test densities similar to the training set are reached with fewer than 100 training densities. A predictor identifies if a test density is within the interpolation region. Via principal component analysis, a projected functional derivative finds highly accurate self-consistent densities. The challenges for application of our method to real electronic structure problems are discussed.},
keywords = {_tablet,2-step model,DFT,dimensionality reduction,KRR,ML,ML-DFA,ML-DFT,ML-ESM,ML-OF,models,orbital-free DFT,original publication,PCA,prediction from density,prediction from density functional,prediction of electron density,prediction of kinetic energy},
- file = {/home/johannes/Nextcloud/Zotero/Snyder et al_2012_Finding Density Functionals with Machine Learning.pdf;/home/johannes/Zotero/storage/6NMNCTQB/Snyder et al. - 2012 - Finding Density Functionals with Machine Learning.tex;/home/johannes/Zotero/storage/TBZPF93I/Snyder et al_2012_Finding Density Functionals with Machine Learning.pdf;/home/johannes/Zotero/storage/RRS5SC4P/PhysRevLett.108.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Snyder et al_2012_Finding Density Functionals with Machine Learning.pdf;/Users/wasmer/Zotero/storage/6NMNCTQB/Snyder et al. - 2012 - Finding Density Functionals with Machine Learning.tex;/Users/wasmer/Zotero/storage/TBZPF93I/Snyder et al_2012_Finding Density Functionals with Machine Learning.pdf;/Users/wasmer/Zotero/storage/RRS5SC4P/PhysRevLett.108.html}
}
@article{soiland-reyesPackagingResearchArtefacts2022,
@@ -6449,7 +7138,23 @@
issue = {Preprint},
langid = {english},
keywords = {FAIR,FDO,JSON-LD,PID,RDM,workflows},
- file = {/home/johannes/Nextcloud/Zotero/Soiland-Reyes et al_2022_Packaging research artefacts with RO-Crate.pdf;/home/johannes/Zotero/storage/X2IWHLC7/ds210053.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Soiland-Reyes et al_2022_Packaging research artefacts with RO-Crate.pdf;/Users/wasmer/Zotero/storage/X2IWHLC7/ds210053.html}
+}
+
+@online{sommer3DSCNewDataset2022,
+ title = {{{3DSC}} - {{A New Dataset}} of {{Superconductors Including Crystal Structures}}},
+ author = {Sommer, Timo and Willa, Roland and Schmalian, Jörg and Friederich, Pascal},
+ date = {2022-12-14},
+ number = {arXiv:2212.06071},
+ eprint = {arXiv:2212.06071},
+ eprinttype = {arxiv},
+ doi = {10.48550/arXiv.2212.06071},
+ url = {http://arxiv.org/abs/2212.06071},
+ urldate = {2023-02-15},
+ abstract = {Data-driven methods, in particular machine learning, can help to speed up the discovery of new materials by finding hidden patterns in existing data and using them to identify promising candidate materials. In the case of superconductors, which are a highly interesting but also a complex class of materials with many relevant applications, the use of data science tools is to date slowed down by a lack of accessible data. In this work, we present a new and publicly available superconductivity dataset ('3DSC'), featuring the critical temperature \$T\_\textbackslash mathrm\{c\}\$ of superconducting materials additionally to tested non-superconductors. In contrast to existing databases such as the SuperCon database which contains information on the chemical composition, the 3DSC is augmented by the approximate three-dimensional crystal structure of each material. We perform a statistical analysis and machine learning experiments to show that access to this structural information improves the prediction of the critical temperature \$T\_\textbackslash mathrm\{c\}\$ of materials. Furthermore, we see the 3DSC not as a finished dataset, but we provide ideas and directions for further research to improve the 3DSC in multiple ways. We are confident that this database will be useful in applying state-of-the-art machine learning methods to eventually find new superconductors.},
+ pubstate = {preprint},
+ keywords = {Database,disordered,disordered SOAP,ensemble learning,magpie,SOAP,superconductor,with-code},
+ file = {/Users/wasmer/Nextcloud/Zotero/Sommer et al_2022_3DSC - A New Dataset of Superconductors Including Crystal Structures.pdf;/Users/wasmer/Zotero/storage/JMMVYJCI/2212.html}
}
@inproceedings{souzaProvenanceDataMachine2019,
@@ -6462,7 +7167,7 @@
abstract = {Machine Learning (ML) has become essential in several industries. In Computational Science and Engineering (CSE), the complexity of the ML lifecycle comes from the large variety of data, scientists' expertise, tools, and workflows. If data are not tracked properly during the lifecycle, it becomes unfeasible to recreate a ML model from scratch or to explain to stackholders how it was created. The main limitation of provenance tracking solutions is that they cannot cope with provenance capture and integration of domain and ML data processed in the multiple workflows in the lifecycle, while keeping the provenance capture overhead low. To handle this problem, in this paper we contribute with a detailed characterization of provenance data in the ML lifecycle in CSE; a new provenance data representation, called PROV-ML, built on top of W3C PROV and ML Schema; and extensions to a system that tracks provenance from multiple workflows to address the characteristics of ML and CSE, and to allow for provenance queries with a standard vocabulary. We show a practical use in a real case in the O\&G industry, along with its evaluation using 239,616 CUDA cores in parallel.},
eventtitle = {2019 {{IEEE}}/{{ACM Workflows}} in {{Support}} of {{Large-Scale Science}} ({{WORKS}})},
keywords = {Computational Science and Engineering,Machine Learning Lifecycle,Workflow Provenance},
- file = {/home/johannes/Nextcloud/Zotero/Souza et al_2019_Provenance Data in the Machine Learning Lifecycle in Computational Science and.pdf;/home/johannes/Zotero/storage/NXAA6T76/8943505.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Souza et al_2019_Provenance Data in the Machine Learning Lifecycle in Computational Science and.pdf;/Users/wasmer/Zotero/storage/NXAA6T76/8943505.html}
}
@book{spaldinMagneticMaterialsFundamentals2010,
@@ -6479,25 +7184,23 @@
abstract = {Magnetic Materials is an excellent introduction to the basics of magnetism, magnetic materials and their applications in modern device technologies. Retaining the concise style of the original, this edition has been thoroughly revised to address significant developments in the field, including the improved understanding of basic magnetic phenomena, new classes of materials, and changes to device paradigms. With homework problems, solutions to selected problems and a detailed list of references, Magnetic Materials continues to be the ideal book for a one-semester course and as a self-study guide for researchers new to the field. New to this edition:Entirely new chapters on Exchange Bias Coupling, Multiferroic and Magnetoelectric Materials, Magnetic InsulatorsRevised throughout, with substantial updates to the chapters on Magnetic Recording and Magnetic Semiconductors, incorporating the latest advances in the fieldNew example problems with worked solutions},
isbn = {978-0-521-88669-7},
keywords = {condensed matter,graduate,magnetism,textbook},
- file = {/home/johannes/Nextcloud/Zotero/Spaldin_2010_Magnetic Materials.pdf;/home/johannes/Zotero/storage/3A42DP7U/4C8C2C5DF32C9E8D528E1E8D26381C1F.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Spaldin_2010_Magnetic Materials.pdf;/Users/wasmer/Zotero/storage/3A42DP7U/4C8C2C5DF32C9E8D528E1E8D26381C1F.html}
}
-@misc{spencerBetterFasterFermionic2020,
+@online{spencerBetterFasterFermionic2020,
title = {Better, {{Faster Fermionic Neural Networks}}},
author = {Spencer, James S. and Pfau, David and Botev, Aleksandar and Foulkes, W. M. C.},
date = {2020-11-13},
number = {arXiv:2011.07125},
- eprint = {2011.07125},
+ eprint = {arXiv:2011.07125},
eprinttype = {arxiv},
- primaryclass = {physics},
- publisher = {{arXiv}},
doi = {10.48550/arXiv.2011.07125},
url = {http://arxiv.org/abs/2011.07125},
urldate = {2022-06-25},
abstract = {The Fermionic Neural Network (FermiNet) is a recently-developed neural network architecture that can be used as a wavefunction Ansatz for many-electron systems, and has already demonstrated high accuracy on small systems. Here we present several improvements to the FermiNet that allow us to set new records for speed and accuracy on challenging systems. We find that increasing the size of the network is sufficient to reach chemical accuracy on atoms as large as argon. Through a combination of implementing FermiNet in JAX and simplifying several parts of the network, we are able to reduce the number of GPU hours needed to train the FermiNet on large systems by an order of magnitude. This enables us to run the FermiNet on the challenging transition of bicyclobutane to butadiene and compare against the PauliNet on the automerization of cyclobutadiene, and we achieve results near the state of the art for both.},
- archiveprefix = {arXiv},
+ pubstate = {preprint},
keywords = {autodiff,DeepMind,FermiNet,JAX,library,MC,ML,ML-ESM,ML-QMBP,NN,PauliNet,prediction of wavefunction,QMC,VMC,with-code},
- file = {/home/johannes/Nextcloud/Zotero/Spencer et al_2020_Better, Faster Fermionic Neural Networks.pdf;/home/johannes/Zotero/storage/SCSQGZ4K/2011.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Spencer et al_2020_Better, Faster Fermionic Neural Networks.pdf;/Users/wasmer/Zotero/storage/SCSQGZ4K/2011.html}
}
@article{spitalerPerspectivesTheoryDefects2018,
@@ -6515,7 +7218,7 @@
abstract = {Our understanding of defects in materials science has changed tremendously over the last century. While one hundred years ago they were often ignored by scientists, nowadays they are in the spotlight of scientific interest and whole branches of technology have emerged from their skillful handling. The first part of this article gives a historical overview and discusses why defects are so important for modern material science. In the second part, we review the treatment of defects in semiconductors. We start by explaining the assumptions and approximations involved in ab-initio calculations and then discuss the treatment of defects in materials. In the third part, we focus on defects in metals. We discuss the theoretical treatment of vacancies in metals starting from experimental findings. The impact of improved theoretical techniques on the predictive power is discussed. This is illustrated with the role of vacancies in intermetallic compounds and random alloys. The last section deals with dislocations.},
langid = {english},
keywords = {defects,First-principles theory,Metals and alloys,review,Semiconductors},
- file = {/home/johannes/Nextcloud/Zotero/Spitaler_Estreicher_2018_Perspectives on the Theory of Defects.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Spitaler_Estreicher_2018_Perspectives on the Theory of Defects.pdf}
}
@article{staackeKernelChargeEquilibration2022,
@@ -6536,7 +7239,7 @@
abstract = {State-of-the-art machine learning (ML) interatomic potentials use local representations of atomic environments to ensure linear scaling and size-extensivity. This implies a neglect of long-range interactions, most prominently related to electrostatics. To overcome this limitation, we herein present a ML framework for predicting charge distributions and their interactions termed kernel charge equilibration (kQEq). This model is based on classical charge equilibration (QEq) models expanded with an environment-dependent electronegativity. In contrast to previously reported neural network models with a similar concept, kQEq takes advantage of the linearity of both QEq and Kernel Ridge Regression to obtain a closed-form linear algebra expression for training the models. Furthermore, we avoid the ambiguity of charge partitioning schemes by using dipole moments as reference data. As a first application, we show that kQEq can be used to generate accurate and highly data-efficient models for molecular dipole moments.},
langid = {english},
keywords = {charge equilibration,charge transfer,electronegativity,kQEq,KRR,ML,molecules},
- file = {/home/johannes/Nextcloud/Zotero/Staacke et al_2022_Kernel charge equilibration.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Staacke et al_2022_Kernel charge equilibration.pdf}
}
@unpublished{steinbachReproducibilityDataScience2022,
@@ -6550,7 +7253,7 @@
urldate = {2022-06-09},
abstract = {Machine Learning is becoming ubiquitous in many scientific domains. However, practitioners struggle to apply every new addition to the Machine Learning market on their data with comparable effects than published. In this talk, I'd like to present recent observations on reproducibility of Machine Learning results and how the community strives to tackle related challenges. Given at https://events.hifis.net/event/426/timetable/},
langid = {english},
- file = {/home/johannes/Nextcloud/Zotero/Steinbach_2022_Reproducibility in Data Science and Machine Learning.pdf;/home/johannes/Zotero/storage/MPANKYEK/1.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Steinbach_2022_Reproducibility in Data Science and Machine Learning.pdf;/Users/wasmer/Zotero/storage/MPANKYEK/1.html}
}
@article{suttonIdentifyingDomainsApplicability2020,
@@ -6570,7 +7273,25 @@
issue = {1},
langid = {english},
keywords = {descriptors,Domains of applicability (DA),ML,models,subgroup discovery},
- file = {/home/johannes/Nextcloud/Zotero/Sutton et al_2020_Identifying domains of applicability of machine learning models for materials.pdf;/home/johannes/Zotero/storage/MPRCKUUI/s41467-020-17112-9.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Sutton et al_2020_Identifying domains of applicability of machine learning models for materials.pdf;/Users/wasmer/Zotero/storage/MPRCKUUI/s41467-020-17112-9.html}
+}
+
+@article{szlachtaAccuracyTransferabilityGaussian2014,
+ title = {Accuracy and Transferability of {{Gaussian}} Approximation Potential Models for Tungsten},
+ author = {Szlachta, Wojciech J. and Bartók, Albert P. and Csányi, Gábor},
+ date = {2014-09-24},
+ journaltitle = {Physical Review B},
+ shortjournal = {Phys. Rev. B},
+ volume = {90},
+ number = {10},
+ pages = {104108},
+ publisher = {{American Physical Society}},
+ doi = {10.1103/PhysRevB.90.104108},
+ url = {https://link.aps.org/doi/10.1103/PhysRevB.90.104108},
+ urldate = {2023-03-12},
+ abstract = {We introduce interatomic potentials for tungsten in the bcc crystal phase and its defects within the Gaussian approximation potential framework, fitted to a database of first-principles density functional theory calculations. We investigate the performance of a sequence of models based on databases of increasing coverage in configuration space and showcase our strategy of choosing representative small unit cells to train models that predict properties observable only using thousands of atoms. The most comprehensive model is then used to calculate properties of the screw dislocation, including its structure, the Peierls barrier and the energetics of the vacancy-dislocation interaction. All software and raw data are available at www.libatoms.org.},
+ keywords = {/unread,AML,database generation,defects,disordered,GAP,GPR,materials,ML,point defects,SOAP,vacancies},
+ file = {/Users/wasmer/Nextcloud/Zotero/Szlachta et al_2014_Accuracy and transferability of Gaussian approximation potential models for.pdf;/Users/wasmer/Nextcloud/Zotero/Szlachta et al_2014_Accuracy and transferability of Gaussian approximation potential models for2.pdf;/Users/wasmer/Zotero/storage/YFHICPLQ/PhysRevB.90.html}
}
@article{talirzMaterialsCloudPlatform2020,
@@ -6591,7 +7312,7 @@
issue = {1},
langid = {english},
annotation = {Bandiera\_abtest: a Cc\_license\_type: cc\_by Cg\_type: Nature Research Journals Primary\_atype: Research Subject\_term: Databases;Materials science Subject\_term\_id: databases;materials-science},
- file = {/home/johannes/Nextcloud/Zotero/Talirz et al_2020_Materials Cloud, a platform for open computational science.pdf;/home/johannes/Zotero/storage/TEZC6LT2/s41597-020-00637-5.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Talirz et al_2020_Materials Cloud, a platform for open computational science.pdf;/Users/wasmer/Zotero/storage/TEZC6LT2/s41597-020-00637-5.html}
}
@unpublished{talirzTrendsAtomisticSimulation2021,
@@ -6600,19 +7321,18 @@
date = {2021-08-27},
eprint = {2108.12350},
eprinttype = {arxiv},
- primaryclass = {cond-mat, physics:physics},
+ eprintclass = {cond-mat, physics:physics},
url = {http://arxiv.org/abs/2108.12350},
urldate = {2021-09-11},
abstract = {Driven by the unprecedented computational power available to scientific research, the use of computers in solid-state physics, chemistry and materials science has been on a continuous rise. This review focuses on the software used for the simulation of matter at the atomic scale. We provide a comprehensive overview of major codes in the field, and analyze how citations to these codes in the academic literature have evolved since 2010. An interactive version of the underlying data set is available at https://atomistic.software .},
- archiveprefix = {arXiv},
keywords = {DFT,DFT codes comparison},
- file = {/home/johannes/Nextcloud/Zotero/Talirz et al_2021_Trends in atomistic simulation software usage.pdf;/home/johannes/Zotero/storage/SCDVYPXG/2108.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Talirz et al_2021_Trends in atomistic simulation software usage.pdf;/Users/wasmer/Zotero/storage/SCDVYPXG/2108.html}
}
@article{tealeDFTExchangeSharing2022,
title = {{{DFT Exchange}}: {{Sharing Perspectives}} on the {{Workhorse}} of {{Quantum Chemistry}} and {{Materials Science}}},
shorttitle = {{{DFT Exchange}}},
- author = {Teale, Andrew and Helgaker, Trygve and Savin, Andreas and Adamo, Carlo and Aradi, Balint and Arbuznikov, Alexei and Ayers, Paul and Baerends, Evert Jan and Barone, Vincenzo and Calaminici, Patrizia and Cances, Eric and Carter, Emily and Chattaraj, Pratim and Chermette, Henry and Ciofini, Ilaria and Crawford, Daniel and Proft, Frank De and Dobson, John and Draxl, Caludia and Frauenheim, Thomas and Fromager, Emmanuel and Fuentealba, Patricio and Gagliardi, Laura and Galli, Giulia and Gao, Jiali and Geerlings, Paul and Gidopoulous, Nikitas and Gill, Peter and Gori-Giorgi, Paola and Gorling, Andreas and Gould, TIm and Grimme, Stefan and Gritsenko, Oleg and Jensen, Hans Jorgen and Johnson, Erin and Jones, Robert and Kaupp, Martin and Koster, Andreas and Kronik, Leeor and Krylov, Anna and Kvaal, Simen and Laestadius, Andre and Levy, Mel and Lewin, Mathieu and Liu, Shubin and Loos, Pierre-Francois and Maitra, Neepa and Neese, Frank and Perdew, John and Pernal, Katarzyna and Pernot, Pascal and Piecuch, Piotr and Rebolini, Elisa and Reining, Lucia and Romaniello, Pina and Ruzsinszky, Adrienn and Salahub, Dennis and Scheffler, Matthias and Schwerdtfeger, Peter and Staroverov, Vicktor and Sun, Jianwei and Tellgren, Erik and Tozer, David and Trickey, Sam and Ullrich, Carsten and Vela, Alberto and Vignale, Giovanni and Wesolowski, Tomasz and Xu, Xin and Yang, Weitao},
+ author = {Teale, Andrew and Helgaker, Trygve and Savin, Andreas and Adamo, Carlo and Aradi, Balint and Arbuznikov, Alexei and Ayers, Paul and Baerends, Evert Jan and Barone, Vincenzo and Calaminici, Patrizia and Cances, Eric and Carter, Emily and Chattaraj, Pratim and Chermette, Henry and Ciofini, Ilaria and Crawford, Daniel and Proft, Frank De and Dobson, John and Draxl, Caludia and Frauenheim, Thomas and Fromager, Emmanuel and Fuentealba, Patricio and Gagliardi, Laura and Galli, Giulia and Gao, Jiali and Geerlings, Paul and Gidopoulous, Nikitas and Gill, Peter and Gori-Giorgi, Paola and Gorling, Andreas and family=Gould, given=Tim, given-i={{Ti}} and Grimme, Stefan and Gritsenko, Oleg and Jensen, Hans Jorgen and Johnson, Erin and Jones, Robert and Kaupp, Martin and Koster, Andreas and Kronik, Leeor and Krylov, Anna and Kvaal, Simen and Laestadius, Andre and Levy, Mel and Lewin, Mathieu and Liu, Shubin and Loos, Pierre-Francois and Maitra, Neepa and Neese, Frank and Perdew, John and Pernal, Katarzyna and Pernot, Pascal and Piecuch, Piotr and Rebolini, Elisa and Reining, Lucia and Romaniello, Pina and Ruzsinszky, Adrienn and Salahub, Dennis and Scheffler, Matthias and Schwerdtfeger, Peter and Staroverov, Vicktor and Sun, Jianwei and Tellgren, Erik and Tozer, David and Trickey, Sam and Ullrich, Carsten and Vela, Alberto and Vignale, Giovanni and Wesolowski, Tomasz and Xu, Xin and Yang, Weitao},
date = {2022-06-17},
doi = {10.26434/chemrxiv-2022-13j2v},
url = {https://chemrxiv.org/engage/chemrxiv/article-details/62974da519595958f0bcc339},
@@ -6620,7 +7340,24 @@
abstract = {In this paper, the history, present status, and future of density-functional theory (DFT) is informally reviewed and discussed by 70 workers in the field, including molecular scientists, materials scientists, method developers and practitioners. The format of the paper is that of a roundtable discussion, in which the participants express and exchange views on DFT in the form of 300 individual contributions, formulated as responses to a preset list of 26 questions. Supported by a bibliography of 776 entries, the paper represents a broad snapshot of DFT, anno 2022.},
langid = {english},
keywords = {DFT,PGI-1/IAS-1,review},
- file = {/home/johannes/Nextcloud/Zotero/Teale et al_2022_DFT Exchange.pdf;/home/johannes/Zotero/storage/JIDYX9CC/62974da519595958f0bcc339.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Teale et al_2022_DFT Exchange.pdf;/Users/wasmer/Zotero/storage/JIDYX9CC/62974da519595958f0bcc339.html}
+}
+
+@online{teufelMEGANMultiExplanationGraph2022,
+ title = {{{MEGAN}}: {{Multi-Explanation Graph Attention Network}}},
+ shorttitle = {{{MEGAN}}},
+ author = {Teufel, Jonas and Torresi, Luca and Reiser, Patrick and Friederich, Pascal},
+ date = {2022-11-23},
+ number = {arXiv:2211.13236},
+ eprint = {arXiv:2211.13236},
+ eprinttype = {arxiv},
+ doi = {10.48550/arXiv.2211.13236},
+ url = {http://arxiv.org/abs/2211.13236},
+ urldate = {2023-03-02},
+ abstract = {Explainable artificial intelligence (XAI) methods are expected to improve trust during human-AI interactions, provide tools for model analysis and extend human understanding of complex problems. Explanation-supervised training allows to improve explanation quality by training self-explaining XAI models on ground truth or human-generated explanations. However, existing explanation methods have limited expressiveness and interoperability due to the fact that only single explanations in form of node and edge importance are generated. To that end we propose the novel multi-explanation graph attention network (MEGAN). Our fully differentiable, attention-based model features multiple explanation channels, which can be chosen independently of the task specifications. We first validate our model on a synthetic graph regression dataset. We show that for the special single explanation case, our model significantly outperforms existing post-hoc and explanation-supervised baseline methods. Furthermore, we demonstrate significant advantages when using two explanations, both in quantitative explanation measures as well as in human interpretability. Finally, we demonstrate our model's capabilities on multiple real-world datasets. We find that our model produces sparse high-fidelity explanations consistent with human intuition about those tasks and at the same time matches state-of-the-art graph neural networks in predictive performance, indicating that explanations and accuracy are not necessarily a trade-off.},
+ pubstate = {preprint},
+ keywords = {/unread,AML,GAT,GNN,materials,ML,XAI},
+ file = {/Users/wasmer/Nextcloud/Zotero/Teufel et al_2022_MEGAN.pdf;/Users/wasmer/Zotero/storage/4ZA3I5AT/2211.html}
}
@thesis{thiessDevelopmentApplicationMassively2011,
@@ -6646,7 +7383,7 @@
isbn = {9783893369065},
langid = {english},
keywords = {juKKR,KKR,KKRnano,PGI-1/IAS-1,thesis},
- file = {/home/johannes/Nextcloud/Zotero/Thieß_2013_Development and application of a massively parallel KKR Green function method.pdf;/home/johannes/Zotero/storage/XL6HDNB2/19395.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Thieß_2013_Development and application of a massively parallel KKR Green function method.pdf;/Users/wasmer/Zotero/storage/XL6HDNB2/19395.html}
}
@article{thiessMassivelyParallelDensity2012,
@@ -6660,7 +7397,7 @@
number = {23},
doi = {10.1103/PhysRevB.85.235103},
keywords = {_tablet,juKKR,KKR,KKRnano,PGI-1/IAS-1},
- file = {/home/johannes/Nextcloud/Zotero/Thiess_2012_Massively parallel density functional calculations for thousands of atoms.pdf;/home/johannes/Zotero/storage/PM97ULPL/PhysRevB.85.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Thiess_2012_Massively parallel density functional calculations for thousands of atoms.pdf;/Users/wasmer/Zotero/storage/PM97ULPL/PhysRevB.85.html}
}
@article{thiessMassivelyParallelDensity2012a,
@@ -6673,7 +7410,7 @@
volume = {85},
number = {23},
doi = {10.1103/PhysRevB.85.235103},
- file = {/home/johannes/Nextcloud/Zotero/Thiess_2012_Massively parallel density functional calculations for thousands of atoms2.pdf;/home/johannes/Zotero/storage/NJSJUCGL/PhysRevB.85.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Thiess_2012_Massively parallel density functional calculations for thousands of atoms2.pdf;/Users/wasmer/Zotero/storage/NJSJUCGL/PhysRevB.85.html}
}
@article{thompsonSpectralNeighborAnalysis2015,
@@ -6691,7 +7428,7 @@
abstract = {We present a new interatomic potential for solids and liquids called Spectral Neighbor Analysis Potential (SNAP). The SNAP potential has a very general form and uses machine-learning techniques to reproduce the energies, forces, and stress tensors of a large set of small configurations of atoms, which are obtained using high-accuracy quantum electronic structure (QM) calculations. The local environment of each atom is characterized by a set of bispectrum components of the local neighbor density projected onto a basis of hyperspherical harmonics in four dimensions. The bispectrum components are the same bond-orientational order parameters employed by the GAP potential [1]. The SNAP potential, unlike GAP, assumes a linear relationship between atom energy and bispectrum components. The linear SNAP coefficients are determined using weighted least-squares linear regression against the full QM training set. This allows the SNAP potential to be fit in a robust, automated manner to large QM data sets using many bispectrum components. The calculation of the bispectrum components and the SNAP potential are implemented in the LAMMPS parallel molecular dynamics code. We demonstrate that a previously unnoticed symmetry property can be exploited to reduce the computational cost of the force calculations by more than one order of magnitude. We present results for a SNAP potential for tantalum, showing that it accurately reproduces a range of commonly calculated properties of both the crystalline solid and the liquid phases. In addition, unlike simpler existing potentials, SNAP correctly predicts the energy barrier for screw dislocation migration in BCC tantalum.},
langid = {english},
keywords = {descriptors,GAP,LAMMPS,MD,ML,PES,SNAP},
- file = {/home/johannes/Nextcloud/Zotero/Thompson et al_2015_Spectral neighbor analysis method for automated generation of quantum-accurate.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Thompson et al_2015_Spectral neighbor analysis method for automated generation of quantum-accurate.pdf}
}
@unpublished{togoSpglibSoftwareLibrary2018,
@@ -6702,15 +7439,14 @@
number = {arXiv:1808.01590},
eprint = {1808.01590},
eprinttype = {arxiv},
- primaryclass = {cond-mat},
+ eprintclass = {cond-mat},
publisher = {{arXiv}},
doi = {10.48550/arXiv.1808.01590},
url = {http://arxiv.org/abs/1808.01590},
urldate = {2022-05-18},
abstract = {A computer algorithm to search crystal symmetries of crystal structures has been implemented in software \$\textbackslash texttt\{spglib\}\$. An iterative algorithm is employed to find a set of space group operations that belongs to any one of space group types by accepting certain amount of distortion for input unit cell structures. The source code is distributed under the BSD 3-Clause License that is a permissive free software licence. Although \$\textbackslash texttt\{spglib\}\$ is a small code, the iteration loops made the source code complicated. The aim of this text is to provide the algorithm details to those people who are interested in inside-\$\textbackslash texttt\{spglib\}\$. This text is written for \$\textbackslash texttt\{spglib\}\$ v1.10.4.},
- archiveprefix = {arXiv},
keywords = {condensed matter,crystal symmetry,library},
- file = {/home/johannes/Nextcloud/Zotero/Togo_Tanaka_2018_$-texttt Spglib $.pdf;/home/johannes/Zotero/storage/67C8WPLU/1808.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Togo_Tanaka_2018_$-texttt Spglib $.pdf;/Users/wasmer/Zotero/storage/67C8WPLU/1808.html}
}
@article{tokuraEmergentFunctionsQuantum2017,
@@ -6731,7 +7467,7 @@
issue = {11},
langid = {english},
annotation = {Bandiera\_abtest: a Cg\_type: Nature Research Journals Primary\_atype: Reviews Subject\_term: Electronic devices;Electronic properties and materials;Ferroelectrics and multiferroics;Superconducting properties and materials;Topological matter Subject\_term\_id: electronic-devices;electronic-properties-and-materials;ferroelectrics-and-multiferroics;superconducting-properties-and-materials;topological-matter},
- file = {/home/johannes/Nextcloud/Zotero/Tokura et al_2017_Emergent functions of quantum materials.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Tokura et al_2017_Emergent functions of quantum materials.pdf}
}
@article{tokuraMagneticSkyrmionMaterials2021,
@@ -6766,7 +7502,7 @@
urldate = {2022-05-13},
abstract = {Solving the coupled-cluster (CC) equations is a cost-prohibitive process that exhibits poor scaling with system size. These equations are solved by determining the set of amplitudes (t) that minimize the system energy with respect to the coupled-cluster equations at the selected level of truncation. Here, a novel approach to predict the converged coupled-cluster singles and doubles (CCSD) amplitudes, thus the coupled-cluster wave function, is explored by using machine learning and electronic structure properties inherent to the MP2 level. Features are collected from quantum chemical data, such as orbital energies, one-electron Hamiltonian, Coulomb, and exchange terms. The data-driven CCSD (DDCCSD) is not an alchemical method because the actual iterative coupled-cluster equations are solved. However, accurate energetics can also be obtained by bypassing solving the CC equations entirely. Our preliminary data show that it is possible to achieve remarkable speedups in solving the CCSD equations, especially when the correct physics are encoded and used for training of machine learning models.},
keywords = {ML,surrogate model},
- file = {/home/johannes/Nextcloud/Zotero/Townsend_Vogiatzis_2019_Data-Driven Acceleration of the Coupled-Cluster Singles and Doubles Iterative.pdf;/home/johannes/Zotero/storage/RVTRBAZI/acs.jpclett.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Townsend_Vogiatzis_2019_Data-Driven Acceleration of the Coupled-Cluster Singles and Doubles Iterative.pdf;/Users/wasmer/Zotero/storage/RVTRBAZI/acs.jpclett.html}
}
@article{uhrinWorkflowsAiiDAEngineering2021,
@@ -6785,7 +7521,7 @@
abstract = {Over the last two decades, the field of computational science has seen a dramatic shift towards incorporating high-throughput computation and big-data analysis as fundamental pillars of the scientific discovery process. This has necessitated the development of tools and techniques to deal with the generation, storage and processing of large amounts of data. In this work we present an in-depth look at the workflow engine powering AiiDA, a widely adopted, highly flexible and database-backed informatics infrastructure with an emphasis on data reproducibility. We detail many of the design choices that were made which were informed by several important goals: the ability to scale from running on individual laptops up to high-performance supercomputers, managing jobs with runtimes spanning from fractions of a second to weeks and scaling up to thousands of jobs concurrently, and all this while maximising robustness. In short, AiiDA aims to be a Swiss army knife for high-throughput computational science. As well as the architecture, we outline important API design choices made to give workflow writers a great deal of liberty whilst guiding them towards writing robust and modular workflows, ultimately enabling them to encode their scientific knowledge to the benefit of the wider scientific community.},
langid = {english},
keywords = {AiiDA,Computational workflows,Data management,Data sharing,Database,Event-based,High-throughput,Provenance,Robust computation},
- file = {/home/johannes/Nextcloud/Zotero/Uhrin et al_2021_Workflows in AiiDA.pdf;/home/johannes/Zotero/storage/KDEGTQ46/S0927025620305772.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Uhrin et al_2021_Workflows in AiiDA.pdf;/Users/wasmer/Zotero/storage/KDEGTQ46/S0927025620305772.html}
}
@online{unitedstatesMaterialsGenomeInitiative,
@@ -6796,7 +7532,7 @@
abstract = {The Materials Genome Initiative is a multi-agency initiative designed to create a new era of policy, resources, and infrastructure that support U.S. institutions in the effort to discover, manufacture, and deploy advanced materials twice as fast, at a fraction of the cost.},
langid = {english},
organization = {{The White House}},
- file = {/home/johannes/Zotero/storage/LEWHVD66/Materials Genome Initiative for Global Competitiveness.pdf;/home/johannes/Zotero/storage/9KCC6KRJ/mgi.html}
+ file = {/Users/wasmer/Zotero/storage/LEWHVD66/Materials Genome Initiative for Global Competitiveness.pdf;/Users/wasmer/Zotero/storage/9KCC6KRJ/mgi.html}
}
@article{unkeMachineLearningForce2021,
@@ -6815,7 +7551,7 @@
urldate = {2021-10-22},
abstract = {In recent years, the use of machine learning (ML) in computational chemistry has enabled numerous advances previously out of reach due to the computational complexity of traditional electronic-structure methods. One of the most promising applications is the construction of ML-based force fields (FFs), with the aim to narrow the gap between the accuracy of ab initio methods and the efficiency of classical FFs. The key idea is to learn the statistical relation between chemical structure and potential energy without relying on a preconceived notion of fixed chemical bonds or knowledge about the relevant interactions. Such universal ML approximations are in principle only limited by the quality and quantity of the reference data used to train them. This review gives an overview of applications of ML-FFs and the chemical insights that can be obtained from them. The core concepts underlying ML-FFs are described in detail, and a step-by-step guide for constructing and testing them from scratch is given. The text concludes with a discussion of the challenges that remain to be overcome by the next generation of ML-FFs.},
keywords = {chemistry,GAP,GDML,GPR,hyperparameters,ML,ML-DFT,ML-ESM,ML-FF,MLP,models,MPNN,NNP,PES,regression,regularization,review,SchNet,sGDML,tutorial},
- file = {/home/johannes/Nextcloud/Zotero/Unke et al_2021_Machine Learning Force Fields2.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Unke et al_2021_Machine Learning Force Fields2.pdf}
}
@inproceedings{unkeSEEquivariantPrediction2021,
@@ -6830,7 +7566,23 @@
urldate = {2022-08-21},
abstract = {Machine learning has enabled the prediction of quantum chemical properties with high accuracy and efficiency, allowing to bypass computationally costly ab initio calculations. Instead of training on a fixed set of properties, more recent approaches attempt to learn the electronic wavefunction (or density) as a central quantity of atomistic systems, from which all other observables can be derived. This is complicated by the fact that wavefunctions transform non-trivially under molecular rotations, which makes them a challenging prediction target. To solve this issue, we introduce general SE(3)-equivariant operations and building blocks for constructing deep learning architectures for geometric point cloud data and apply them to reconstruct wavefunctions of atomistic systems with unprecedented accuracy. Our model achieves speedups of over three orders of magnitude compared to ab initio methods and reduces prediction errors by up to two orders of magnitude compared to the previous state-of-the-art. This accuracy makes it possible to derive properties such as energies and forces directly from the wavefunction in an end-to-end manner. We demonstrate the potential of our approach in a transfer learning application, where a model trained on low accuracy reference wavefunctions implicitly learns to correct for electronic many-body interactions from observables computed at a higher level of theory. Such machine-learned wavefunction surrogates pave the way towards novel semi-empirical methods, offering resolution at an electronic level while drastically decreasing computational cost. Additionally, the predicted wavefunctions can serve as initial guess in conventional ab initio methods, decreasing the number of iterations required to arrive at a converged solution, thus leading to significant speedups without any loss of accuracy or robustness. While we focus on physics applications in this contribution, the proposed equivariant framework for deep learning on point clouds is promising also beyond, say, in computer vision or graphics.},
keywords = {_tablet,EGNN,equivariant,initial guess,ML-ESM,original publication,PhiSNet,prediction of electron density,prediction of wavefunction,SchNOrb,with-review},
- file = {/home/johannes/Nextcloud/Zotero/Unke et al_2021_SE(3)-equivariant prediction of molecular wavefunctions and electronic densities.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Unke et al_2021_SE(3)-equivariant prediction of molecular wavefunctions and electronic densities.pdf}
+}
+
+@online{vanderoordHyperactiveLearningHAL2022,
+ title = {Hyperactive {{Learning}} ({{HAL}}) for {{Data-Driven Interatomic Potentials}}},
+ author = {family=Oord, given=Cas, prefix=van der, useprefix=true and Sachs, Matthias and Kovács, Dávid Péter and Ortner, Christoph and Csányi, Gábor},
+ date = {2022-11-07},
+ number = {arXiv:2210.04225},
+ eprint = {arXiv:2210.04225},
+ eprinttype = {arxiv},
+ doi = {10.48550/arXiv.2210.04225},
+ url = {http://arxiv.org/abs/2210.04225},
+ urldate = {2023-02-05},
+ abstract = {Data-driven interatomic potentials have emerged as a powerful class of surrogate models for \{\textbackslash it ab initio\} potential energy surfaces that are able to reliably predict macroscopic properties with experimental accuracy. In generating accurate and transferable potentials the most time-consuming and arguably most important task is generating the training set, which still requires significant expert user input. To accelerate this process, this work presents \textbackslash text\{\textbackslash it hyperactive learning\} (HAL), a framework for formulating an accelerated sampling algorithm specifically for the task of training database generation. The key idea is to start from a physically motivated sampler (e.g., molecular dynamics) and add a biasing term that drives the system towards high uncertainty and thus to unseen training configurations. Building on this framework, general protocols for building training databases for alloys and polymers leveraging the HAL framework will be presented. For alloys, ACE potentials for AlSi10 are created by fitting to a minimal HAL-generated database containing 88 configurations (32 atoms each) with fast evaluation times of {$<$}100 microsecond/atom/cpu-core. These potentials are demonstrated to predict the melting temperature with excellent accuracy. For polymers, a HAL database is built using ACE, able to determine the density of a long polyethylene glycol (PEG) polymer formed of 200 monomer units with experimental accuracy by only fitting to small isolated PEG polymers with sizes ranging from 2 to 32.},
+ pubstate = {preprint},
+ keywords = {/unread,Physics - Computational Physics,Statistics - Machine Learning},
+ file = {/Users/wasmer/Zotero/storage/4S2GHGVG/van der Oord et al. - 2022 - Hyperactive Learning (HAL) for Data-Driven Interat.pdf;/Users/wasmer/Zotero/storage/YJBLUYLE/2210.html}
}
@article{vannoordenTop100Papers2014,
@@ -6848,27 +7600,25 @@
langid = {english},
keywords = {DFT,for introductions,literature analysis,Nature,popular science},
annotation = {Cg\_type: Nature News},
- file = {/home/johannes/Zotero/storage/5L7C8V9H/the-top-100-papers-1.html}
+ file = {/Users/wasmer/Zotero/storage/5L7C8V9H/the-top-100-papers-1.html}
}
@online{vanrossumPEPStyleGuide,
title = {{{PEP}} 8 -- {{Style Guide}} for {{Python Code}}},
- author = {van Rossum, Guido and Warsaw, Barry and Coghlan, Nick},
- options = {useprefix=true},
+ author = {family=Rossum, given=Guido, prefix=van, useprefix=true and Warsaw, Barry and Coghlan, Nick},
url = {https://www.python.org/dev/peps/pep-0008/},
urldate = {2021-09-23},
abstract = {The official home of the Python Programming Language},
langid = {english},
organization = {{Python.org}},
keywords = {coding style guide,PEP,Python,software engineering},
- file = {/home/johannes/Zotero/storage/A4H9CLJ5/pep-0008.html}
+ file = {/Users/wasmer/Zotero/storage/A4H9CLJ5/pep-0008.html}
}
@article{vanschorenOpenMLNetworkedScience2014,
title = {{{OpenML}}: Networked Science in Machine Learning},
shorttitle = {{{OpenML}}},
- author = {Vanschoren, Joaquin and van Rijn, Jan N. and Bischl, Bernd and Torgo, Luis},
- options = {useprefix=true},
+ author = {Vanschoren, Joaquin and family=Rijn, given=Jan N., prefix=van, useprefix=true and Bischl, Bernd and Torgo, Luis},
date = {2014-06-16},
journaltitle = {ACM SIGKDD Explorations Newsletter},
shortjournal = {SIGKDD Explor. Newsl.},
@@ -6882,9 +7632,8 @@
url = {http://arxiv.org/abs/1407.7722},
urldate = {2022-01-02},
abstract = {Many sciences have made significant breakthroughs by adopting online tools that help organize, structure and mine information that is too detailed to be printed in journals. In this paper, we introduce OpenML, a place for machine learning researchers to share and organize data in fine detail, so that they can work more effectively, be more visible, and collaborate with others to tackle harder problems. We discuss how OpenML relates to other examples of networked science and what benefits it brings for machine learning research, individual scientists, as well as students and practitioners.},
- archiveprefix = {arXiv},
keywords = {Computer Science - Computers and Society,Computer Science - Machine Learning},
- file = {/home/johannes/Nextcloud/Zotero/Vanschoren et al_2014_OpenML.pdf;/home/johannes/Zotero/storage/YFGEM7US/1407.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Vanschoren et al_2014_OpenML.pdf;/Users/wasmer/Zotero/storage/YFGEM7US/1407.html}
}
@inproceedings{vartakModelDBSystemMachine2016,
@@ -6902,7 +7651,7 @@
urldate = {2021-10-23},
abstract = {Building a machine learning model is an iterative process. A data scientist will build many tens to hundreds of models before arriving at one that meets some acceptance criteria (e.g. AUC cutoff, accuracy threshold). However, the current style of model building is ad-hoc and there is no practical way for a data scientist to manage models that are built over time. As a result, the data scientist must attempt to "remember" previously constructed models and insights obtained from them. This task is challenging for more than a handful of models and can hamper the process of sensemaking. Without a means to manage models, there is no easy way for a data scientist to answer questions such as "Which models were built using an incorrect feature?", "Which model performed best on American customers?" or "How did the two top models compare?" In this paper, we describe our ongoing work on ModelDB, a novel end-to-end system for the management of machine learning models. ModelDB clients automatically track machine learning models in their native environments (e.g. scikit-learn, spark.ml), the ModelDB backend introduces a common layer of abstractions to represent models and pipelines, and the ModelDB frontend allows visual exploration and analyses of models via a web-based interface.},
isbn = {978-1-4503-4207-0},
- file = {/home/johannes/Nextcloud/Zotero/Vartak et al_2016_Mspan class=smallcaps smallerCapitalodel-spanDB.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Vartak et al_2016_Mspan class=smallcaps smallerCapitalodel-spanDB.pdf}
}
@article{vedmedenko2020MagnetismRoadmap2020,
@@ -6922,7 +7671,7 @@
abstract = {Andreas Berger CICnanoGUNE BRTA Following the success and relevance of the 2014 and 2017 Magnetism Roadmap articles, this 2020 Magnetism Roadmap edition takes yet another timely look at newly relevant and highly active areas in magnetism research. The overall layout of this article is unchanged, given that it has proved the most appropriate way to convey the most relevant aspects of today’s magnetism research in a wide variety of sub-fields to a broad readership. A different group of experts has again been selected for this article, representing both the breadth of new research areas, and the desire to incorporate different voices and viewpoints. The latter is especially relevant for thistype of article, in which one’s field of expertise has to be accommodated on two printed pages only, so that personal selection preferences are naturally rather more visible than in other types of articles. Most importantly, the very relevant advances in the field of magnetism research in recent years make the publication of yet another Magnetism Roadmap a very sensible and timely endeavour, allowing its authors and readers to take another broad-based, but concise look at the most significant developments in magnetism, their precise status, their challenges, and their anticipated future developments. While many of the contributions in this 2020 Magnetism Roadmap edition have significant associations with different aspects of magnetism, the general layout can nonetheless be classified in terms of three main themes: (i) phenomena, (ii) materials and characterization, and (iii) applications and devices. While these categories are unsurprisingly rather similar to the 2017 Roadmap, the order is different, in that the 2020 Roadmap considers phenomena first, even if their occurrences are naturally very difficult to separate from the materials exhibiting such phenomena. Nonetheless, the specifically selected topics seemed to be best displayed in the order presented here, in particular, because many of the phenomena or geometries discussed in (i) can be found or designed into a large variety of materials, so that the progression of the article embarks from more general concepts to more specific classes of materials in the selected order. Given that applications and devices are based on both phenomena and materials, it seemed most appropriate to close the article with the application and devices section (iii) once again. The 2020 Magnetism Roadmap article contains 14 sections, all of which were written by individual authors and experts, specifically addressing a subject in terms of its status, advances, challenges and perspectives in just two pages. Evidently, this two-page format limits the depth to which each subject can be described. Nonetheless, the most relevant and key aspects of each field are touched upon, which enables the Roadmap as whole to give its readership an initial overview of and outlook into a wide variety of topics and fields in a fairly condensed format. Correspondingly, the Roadmap pursues the goal of giving each reader a brief reference frame of relevant and current topics in modern applied magnetism research, even if not all sub-fields can be represented here. The first block of this 2020 Magnetism Roadmap, which is focussed on (i) phenomena, contains five contributions, which address the areas of interfacial Dzyaloshinskii–Moriya interactions, and two-dimensional and curvilinear magnetism, as well as spin-orbit torque phenomena and all optical magnetization reversal. All of these contributions describe cutting edge aspects of rather fundamental physical processes and properties, associated with new and improved magnetic materials’ properties, together with potential developments in terms of future devices and technology. As such, they form part of a widening magnetism ‘phenomena reservoir’ for utilization in applied magnetism and related device technology. The final block (iii) of this article focuses on such applications and device-related fields in four contributions relating to currently active areas of research, which are of course utilizing magnetic phenomena to enable specific functions. These contributions highlight the role of magnetism or spintronics in the field of neuromorphic and reservoir computing, terahertz technology, and domain wall-based logic. One aspect common to all of these application-related contributions is that they are not yet being utilized in commercially available technology; it is currently still an open question, whether or not such technological applications will be magnetism-based at all in the future, or if other types of materials and phenomena will yet outperform magnetism. This last point is actually a very good indication of the vibrancy of applied magnetism research today, given that it demonstrates that magnetism research is able to venture into novel application fields, based upon its portfolio of phenomena, effects and materials. This materials portfolio in particular defines the central block (ii) of this article, with its five contributions interconnecting phenomena with devices, for which materials and the characterization of their properties is the decisive discriminator between purely academically interesting aspects and the true viability of real-life devices, because only available materials and their associated fabrication and characterization methods permit reliable technological implementation. These five contributions specifically address magnetic films and multiferroic heterostructures for the purpose of spin electronic utilization, multi-scale materials modelling, and magnetic materials design based upon machine-learning, as well as materials characterization via polarized neutron measurements. As such, these contributions illustrate the balanced relevance of research into experimental and modelling magnetic materials, as well the importance of sophisticated characterization methods that allow for an ever-more refined understanding of materials. As a combined and integrated article, this 2020 Magnetism Roadmap is intended to be a reference point for current, novel and emerging research directions in modern magnetism, just as its 2014 and 2017 predecessors have been in previous years.},
langid = {english},
keywords = {magnetism,physics,review,roadmap},
- file = {/home/johannes/Nextcloud/Zotero/Vedmedenko et al_2020_The 2020 magnetism roadmap.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Vedmedenko et al_2020_The 2020 magnetism roadmap.pdf}
}
@article{vojvodicExploringLimitsLowpressure2014,
@@ -6941,7 +7690,7 @@
abstract = {The Haber–Bosch process for ammonia synthesis has been suggested to be the most important invention of the 20th century, and called the ‘Bellwether reaction in heterogeneous catalysis’. We examine the catalyst requirements for a new low-pressure, low-temperature synthesis process. We show that the absence of such a process for conventional transition metal catalysts can be understood as a consequence of a scaling relation between the activation energy for N2 dissociation and N adsorption energy found at the surface of these materials. A better catalyst cannot obey this scaling relation. We define the ideal scaling relation characterizing the most active catalyst possible, and show that it is theoretically possible to have a low pressure, low-temperature Haber–Bosch process. The challenge is to find new classes of catalyst materials with properties approaching the ideal, and we discuss the possibility that transition metal compounds have such properties.},
langid = {english},
keywords = {applications of DFT,DFT,master-thesis},
- file = {/home/johannes/Zotero/storage/86UUIB6S/S000926141400147X.html}
+ file = {/Users/wasmer/Zotero/storage/86UUIB6S/S000926141400147X.html}
}
@article{voskoAccurateSpindependentElectron1980,
@@ -6960,7 +7709,7 @@
url = {https://cdnsciencepub.com/doi/abs/10.1139/p80-159},
urldate = {2021-10-18},
keywords = {DFT,LDA,LSDA,original publication,xc functional},
- file = {/home/johannes/Nextcloud/Zotero/Vosko et al_1980_Accurate spin-dependent electron liquid correlation energies for local spin.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Vosko et al_1980_Accurate spin-dependent electron liquid correlation energies for local spin.pdf}
}
@article{vuUnderstandingKernelRidge2015,
@@ -6979,8 +7728,7 @@
abstract = {Accurate approximations to density functionals have recently been obtained via machine learning (ML). By applying ML to a simple function of one variable without any random sampling, we extract the qualitative dependence of errors on hyperparameters. We find universal features of the behavior in extreme limits, including both very small and very large length scales, and the noise-free limit. We show how such features arise in ML models of density functionals. © 2015 Wiley Periodicals, Inc.},
langid = {english},
keywords = {_tablet,density functional theory,DFT,extreme behaviors,hyperparameters optimization,KRR,machine learning,ML,models,noise-free curve,tutorial},
- annotation = {\_eprint: https://onlinelibrary.wiley.com/doi/pdf/10.1002/qua.24939},
- file = {/home/johannes/Nextcloud/Zotero/false;/home/johannes/Nextcloud/Zotero/false;/home/johannes/Nextcloud/Zotero/Vu et al_2015_Understanding kernel ridge regression.pdf;/home/johannes/Zotero/storage/5INUIEQC/qua.html}
+ file = {/home/johannes/Nextcloud/Zotero/false;/home/johannes/Nextcloud/Zotero/false;/Users/wasmer/Nextcloud/Zotero/Vu et al_2015_Understanding kernel ridge regression.pdf;/Users/wasmer/Zotero/storage/5INUIEQC/qua.html}
}
@book{vvedenskySymmetryGroupsRepresentations2010,
@@ -6997,22 +7745,20 @@
annotation = {OCLC: 633422775}
}
-@misc{wangGraphNetsPartial2019,
+@online{wangGraphNetsPartial2019,
title = {Graph {{Nets}} for {{Partial Charge Prediction}}},
author = {Wang, Yuanqing and Fass, Josh and Stern, Chaya D. and Luo, Kun and Chodera, John},
date = {2019-09-17},
number = {arXiv:1909.07903},
- eprint = {1909.07903},
+ eprint = {arXiv:1909.07903},
eprinttype = {arxiv},
- primaryclass = {physics},
- publisher = {{arXiv}},
doi = {10.48550/arXiv.1909.07903},
url = {http://arxiv.org/abs/1909.07903},
urldate = {2022-09-27},
abstract = {Atomic partial charges are crucial parameters for Molecular Dynamics (MD) simulations, molecular mechanics calculations, and virtual screening, as they determine the electrostatic contributions to interaction energies. Current methods for calculating partial charges, however, are either slow and scale poorly with molecular size (quantum chemical methods) or unreliable (empirical methods). Here, we present a new charge derivation method based on Graph Nets---a set of update and aggregate functions that operate on molecular topologies and propagate information thereon---that could approximate charges derived from Density Functional Theory (DFT) calculations with high accuracy and an over 500-fold speed up.},
- archiveprefix = {arXiv},
+ pubstate = {preprint},
keywords = {GCN,GNN,molecules,prediction of partial charge},
- file = {/home/johannes/Nextcloud/Zotero/Wang et al_2019_Graph Nets for Partial Charge Prediction.pdf;/home/johannes/Zotero/storage/5MD2WVP3/1909.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Wang et al_2019_Graph Nets for Partial Charge Prediction.pdf;/Users/wasmer/Zotero/storage/5MD2WVP3/1909.html}
}
@article{wangMachineLearningMaterials2020,
@@ -7032,7 +7778,7 @@
urldate = {2021-05-13},
abstract = {This Methods/Protocols article is intended for materials scientists interested in performing machine learning-centered research. We cover broad guidelines and best practices regarding the obtaining and treatment of data, feature engineering, model training, validation, evaluation and comparison, popular repositories for materials data and benchmarking data sets, model and architecture sharing, and finally publication. In addition, we include interactive Jupyter notebooks with example Python code to demonstrate some of the concepts, workflows, and best practices discussed. Overall, the data-driven methods and machine learning workflows and considerations are presented in a simple way, allowing interested readers to more intelligently guide their machine learning research using the suggested references, best practices, and their own materials domain expertise.},
keywords = {ML,models,notebooks,review,with-code},
- file = {/home/johannes/Nextcloud/Zotero/Wang et al_2020_Machine Learning for Materials Scientists.pdf;/home/johannes/Zotero/storage/PY7PFU35/acs.chemmater.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Wang et al_2020_Machine Learning for Materials Scientists.pdf;/Users/wasmer/Zotero/storage/PY7PFU35/acs.chemmater.html}
}
@article{wangTopologicalStatesCondensed2017,
@@ -7053,7 +7799,7 @@
issue = {11},
langid = {english},
annotation = {Bandiera\_abtest: a Cg\_type: Nature Research Journals Primary\_atype: Reviews Subject\_term: Electronic properties and materials;Quantum Hall;Superconducting properties and materials;Topological matter Subject\_term\_id: electronic-properties-and-materials;quantum-hall;superconducting-properties-and-materials;topological-matter},
- file = {/home/johannes/Nextcloud/Zotero/Wang_Zhang_2017_Topological states of condensed matter.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Wang_Zhang_2017_Topological states of condensed matter.pdf}
}
@article{wangWillAnyCrap2020,
@@ -7071,7 +7817,7 @@
url = {https://doi.org/10.1021/acsnano.9b00184},
urldate = {2022-10-21},
keywords = {physics,rec-by-ghosh,skeptics},
- file = {/home/johannes/Nextcloud/Zotero/Wang et al_2020_Will Any Crap We Put into Graphene Increase Its Electrocatalytic Effect.pdf;/home/johannes/Zotero/storage/5YHQ95UH/acsnano.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Wang et al_2020_Will Any Crap We Put into Graphene Increase Its Electrocatalytic Effect.pdf;/Users/wasmer/Zotero/storage/5YHQ95UH/acsnano.html}
}
@article{wardGeneralpurposeMachineLearning2016,
@@ -7091,7 +7837,7 @@
issue = {1},
langid = {english},
keywords = {compositional descriptors,descriptors,feature engineering,library,magpie,ML,models,rec-by-ruess},
- file = {/home/johannes/Nextcloud/Zotero/Ward et al_2016_A general-purpose machine learning framework for predicting properties of.pdf;/home/johannes/Zotero/storage/39BDW9ZZ/npjcompumats201628.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Ward et al_2016_A general-purpose machine learning framework for predicting properties of.pdf;/Users/wasmer/Zotero/storage/39BDW9ZZ/npjcompumats201628.html}
}
@article{wardMatminerOpenSource2018,
@@ -7110,7 +7856,7 @@
abstract = {As materials data sets grow in size and scope, the role of data mining and statistical learning methods to analyze these materials data sets and build predictive models is becoming more important. This manuscript introduces matminer, an open-source, Python-based software platform to facilitate data-driven methods of analyzing and predicting materials properties. Matminer provides modules for retrieving large data sets from external databases such as the Materials Project, Citrination, Materials Data Facility, and Materials Platform for Data Science. It also provides implementations for an extensive library of feature extraction routines developed by the materials community, with 47 featurization classes that can generate thousands of individual descriptors and combine them into mathematical functions. Finally, matminer provides a visualization module for producing interactive, shareable plots. These functions are designed in a way that integrates closely with machine learning and data analysis packages already developed and in use by the Python data science community. We explain the structure and logic of matminer, provide a description of its various modules, and showcase several examples of how matminer can be used to collect data, reproduce data mining studies reported in the literature, and test new methodologies.},
langid = {english},
keywords = {Data mining,Machine learning,Materials informatics,Open source software},
- file = {/home/johannes/Nextcloud/Zotero/Ward et al_2018_Matminer.pdf;/home/johannes/Zotero/storage/P7KMD3SZ/S0927025618303252.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Ward et al_2018_Matminer.pdf;/Users/wasmer/Zotero/storage/P7KMD3SZ/S0927025618303252.html}
}
@article{waroquiersStatisticalAnalysisCoordination2017,
@@ -7129,7 +7875,7 @@
urldate = {2021-07-20},
abstract = {Coordination or local environments (e.g., tetrahedra and octahedra) are powerful descriptors of the crystalline structure of materials. These structural descriptors are essential to the understanding of crystal chemistry and the design of new materials. However, extensive statistics on the occurrence of local environment are not available even on common chemistries such as oxides. Here, we present the first large-scale statistical analysis of the coordination environments of cations in oxides using a large set of experimentally observed compounds (about 8000). Using a newly developed method, we provide the distribution of local environment for each cation in oxides. We discuss our results highlighting previously known trends and unexpected coordination environments, as well as compounds presenting very rare coordinations. Our work complements the know-how of the solid state chemist with a statistically sound analysis and paves the way for further data mining efforts linking, for instance, coordination environments to materials properties.},
keywords = {ChemEnv,continuous symmetry measure,coordination environments,descriptors,library,pymatgen,rec-by-kovacik,voronoi analysis,voronoi tessellation,with-code},
- file = {/home/johannes/Nextcloud/Zotero/Waroquiers et al_2017_Statistical Analysis of Coordination Environments in Oxides.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Waroquiers et al_2017_Statistical Analysis of Coordination Environments in Oxides.pdf}
}
@software{wasmerAiidajutools2022,
@@ -7154,7 +7900,7 @@
eventtitle = {{{DPG SKM21}}},
langid = {english},
venue = {{online}},
- file = {/home/johannes/Zotero/storage/RJIXGQPU/901958.html}
+ file = {/Users/wasmer/Zotero/storage/RJIXGQPU/901958.html}
}
@thesis{wasmerDevelopmentSurrogateMachine2021,
@@ -7167,8 +7913,8 @@
urldate = {2022-08-08},
langid = {english},
pagetotal = {99},
- keywords = {_tablet,master-thesis,PGI-1/IAS-1,thesis},
- file = {/home/johannes/Nextcloud/Zotero/false;/home/johannes/Nextcloud/Zotero/Wasmer_2021_Development of a surrogate machine learning model for the acceleration of.pdf;/home/johannes/Zotero/storage/AC483X2N/master-thesis.html}
+ keywords = {_tablet,AiiDA,aiida-kkr,AML,Coulomb matrix,descriptor comparison,impurity embedding,juKKR,KKR,master-thesis,ML,PGI-1/IAS-1,SOAP,thesis},
+ file = {/home/johannes/Nextcloud/Zotero/false;/Users/wasmer/Nextcloud/Zotero/Wasmer_2021_Development of a surrogate machine learning model for the acceleration of.pdf;/Users/wasmer/Zotero/storage/AC483X2N/master-thesis.html}
}
@article{weinertSolutionPoissonEquation1981,
@@ -7187,21 +7933,22 @@
url = {https://aip.scitation.org/doi/10.1063/1.524800},
urldate = {2022-10-01},
keywords = {DFT,FLEUR,Poisson equation,potential},
- file = {/home/johannes/Nextcloud/Zotero/Weinert_1981_Solution of Poisson’s equation.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Weinert_1981_Solution of Poisson’s equation.pdf}
}
-@misc{wellawattePerspectiveExplanationsMolecular2022,
+@online{wellawattePerspectiveExplanationsMolecular2022,
title = {A {{Perspective}} on {{Explanations}} of {{Molecular Prediction Models}}},
author = {Wellawatte, Geemi P. and Gandhi, Heta A. and Seshadri, Aditi and White, Andrew D.},
date = {2022-12-09},
- publisher = {{ChemRxiv}},
+ eprinttype = {ChemRxiv},
doi = {10.26434/chemrxiv-2022-qfv02},
url = {https://chemrxiv.org/engage/chemrxiv/article-details/639222a114d92d7cd6a65e90},
urldate = {2022-12-29},
abstract = {Chemists can be skeptical in using deep learning (DL) in decision making, due to the lack of interpretability in "black-box" models. Explainable artificial intelligence (XAI) is a branch of AI which addresses this drawback by providing tools to interpret DL models and their predictions. We review the principles of XAI in the domain of chemistry and emerging methods for creating and evaluating explanations. Then we focus methods developed by our group and their application to predicting solubility, blood-brain barrier permeability, and the scent of molecules. We show that XAI methods like chemical counterfactuals and descriptor explanations can both explain DL predictions and give insight into structure-property relationships. Finally, we discuss how a two step process of highly accurate black-box modeling and then creating explanations gives both highly accurate predictions and clear structure-property relationships.},
langid = {english},
- keywords = {counterfactual explanation,Deep learning,GNN,molecules,XAI},
- file = {/home/johannes/Nextcloud/Zotero/Wellawatte et al_2022_A Perspective on Explanations of Molecular Prediction Models.pdf}
+ pubstate = {preprint},
+ keywords = {_tablet,counterfactual explanation,Deep learning,GNN,molecules,XAI},
+ file = {/Users/wasmer/Nextcloud/Zotero/Wellawatte et al_2022_A Perspective on Explanations of Molecular Prediction Models.pdf}
}
@article{westermayrMachineLearningElectronically2021,
@@ -7220,7 +7967,7 @@
urldate = {2021-12-14},
abstract = {Electronically excited states of molecules are at the heart of photochemistry, photophysics, as well as photobiology and also play a role in material science. Their theoretical description requires highly accurate quantum chemical calculations, which are computationally expensive. In this review, we focus on not only how machine learning is employed to speed up such excited-state simulations but also how this branch of artificial intelligence can be used to advance this exciting research field in all its aspects. Discussed applications of machine learning for excited states include excited-state dynamics simulations, static calculations of absorption spectra, as well as many others. In order to put these studies into context, we discuss the promises and pitfalls of the involved machine learning techniques. Since the latter are mostly based on quantum chemistry calculations, we also provide a short introduction into excited-state electronic structure methods and approaches for nonadiabatic dynamics simulations and describe tricks and problems when using them in machine learning for excited states of molecules.},
keywords = {DFT,excited states,MD,ML,molecules,review,TDDFT},
- file = {/home/johannes/Nextcloud/Zotero/Westermayr_Marquetand_2021_Machine Learning for Electronically Excited States of Molecules.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Westermayr_Marquetand_2021_Machine Learning for Electronically Excited States of Molecules.pdf}
}
@unpublished{westermayrPerspectiveIntegratingMachine2021,
@@ -7229,13 +7976,12 @@
date = {2021-04-20},
eprint = {2102.08435},
eprinttype = {arxiv},
- primaryclass = {physics},
+ eprintclass = {physics},
url = {http://arxiv.org/abs/2102.08435},
urldate = {2021-05-13},
abstract = {Machine learning (ML) methods are being used in almost every conceivable area of electronic structure theory and molecular simulation. In particular, ML has become firmly established in the construction of high-dimensional interatomic potentials. Not a day goes by without another proof of principle being published on how ML methods can represent and predict quantum mechanical properties -- be they observable, such as molecular polarizabilities, or not, such as atomic charges. As ML is becoming pervasive in electronic structure theory and molecular simulation, we provide an overview of how atomistic computational modeling is being transformed by the incorporation of ML approaches. From the perspective of the practitioner in the field, we assess how common workflows to predict structure, dynamics, and spectroscopy are affected by ML. Finally, we discuss how a tighter and lasting integration of ML methods with computational chemistry and materials science can be achieved and what it will mean for research practice, software development, and postgraduate training.},
- archiveprefix = {arXiv},
keywords = {ML,physics,review},
- file = {/home/johannes/Nextcloud/Zotero/Westermayr et al_2021_Perspective on integrating machine learning into computational chemistry and.pdf;/home/johannes/Nextcloud/Zotero/Westermayr et al_2021_Perspective on integrating machine learning into computational chemistry and2.pdf;/home/johannes/Zotero/storage/FHJLNQAU/2102.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Westermayr et al_2021_Perspective on integrating machine learning into computational chemistry and.pdf;/Users/wasmer/Nextcloud/Zotero/Westermayr et al_2021_Perspective on integrating machine learning into computational chemistry and2.pdf;/Users/wasmer/Zotero/storage/FHJLNQAU/2102.html}
}
@article{whiteDeepLearningMolecules2021,
@@ -7254,7 +8000,7 @@
issue = {1},
langid = {english},
keywords = {_tablet,book,GNN,ML-DFT,ML-ESM,MLP,MPNN,tutorial},
- file = {/home/johannes/Nextcloud/Zotero/White_2021_Deep Learning for Molecules and Materials.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/White_2021_Deep Learning for Molecules and Materials.pdf}
}
@article{wiederCompactReviewMolecular2020,
@@ -7272,7 +8018,7 @@
abstract = {As graph neural networks are becoming more and more powerful and useful in the field of drug discovery, many pharmaceutical companies are getting interested in utilizing these methods for their own in-house frameworks. This is especially compelling for tasks such as the prediction of molecular properties which is often one of the most crucial tasks in computer-aided drug discovery workflows. The immense hype surrounding these kinds of algorithms has led to the development of many different types of promising architectures and in this review we try to structure this highly dynamic field of AI-research by collecting and classifying 80 GNNs that have been used to predict more than 20 molecular properties using 48 different datasets.},
langid = {english},
keywords = {_tablet,chemistry,GCN,GNN,molecules,review},
- file = {/home/johannes/Nextcloud/Zotero/Wieder et al_2020_A compact review of molecular property prediction with graph neural networks.pdf;/home/johannes/Zotero/storage/KHCYV2ZB/S1740674920300305.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Wieder et al_2020_A compact review of molecular property prediction with graph neural networks.pdf;/Users/wasmer/Zotero/storage/KHCYV2ZB/S1740674920300305.html}
}
@article{wilkinsAccurateMolecularPolarizabilities2019,
@@ -7288,13 +8034,12 @@
url = {https://www.pnas.org/doi/10.1073/pnas.1816132116},
urldate = {2022-08-16},
keywords = {ML-ESM,SA-GPR},
- file = {/home/johannes/Nextcloud/Zotero/Wilkins et al_2019_Accurate molecular polarizabilities with coupled cluster theory and machine.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Wilkins et al_2019_Accurate molecular polarizabilities with coupled cluster theory and machine.pdf}
}
@article{wilkinsonFAIRGuidingPrinciples2016,
title = {The {{FAIR Guiding Principles}} for Scientific Data Management and Stewardship},
- author = {Wilkinson, Mark D. and Dumontier, Michel and Aalbersberg, IJsbrand Jan and Appleton, Gabrielle and Axton, Myles and Baak, Arie and Blomberg, Niklas and Boiten, Jan-Willem and da Silva Santos, Luiz Bonino and Bourne, Philip E. and Bouwman, Jildau and Brookes, Anthony J. and Clark, Tim and Crosas, Mercè and Dillo, Ingrid and Dumon, Olivier and Edmunds, Scott and Evelo, Chris T. and Finkers, Richard and Gonzalez-Beltran, Alejandra and Gray, Alasdair J. G. and Groth, Paul and Goble, Carole and Grethe, Jeffrey S. and Heringa, Jaap and ’t Hoen, Peter A. C. and Hooft, Rob and Kuhn, Tobias and Kok, Ruben and Kok, Joost and Lusher, Scott J. and Martone, Maryann E. and Mons, Albert and Packer, Abel L. and Persson, Bengt and Rocca-Serra, Philippe and Roos, Marco and van Schaik, Rene and Sansone, Susanna-Assunta and Schultes, Erik and Sengstag, Thierry and Slater, Ted and Strawn, George and Swertz, Morris A. and Thompson, Mark and van der Lei, Johan and van Mulligen, Erik and Velterop, Jan and Waagmeester, Andra and Wittenburg, Peter and Wolstencroft, Katherine and Zhao, Jun and Mons, Barend},
- options = {useprefix=true},
+ author = {Wilkinson, Mark D. and Dumontier, Michel and family=Aalbersberg, given=IJsbrand Jan, given-i={{IJ}}J and Appleton, Gabrielle and Axton, Myles and Baak, Arie and Blomberg, Niklas and Boiten, Jan-Willem and family=Silva Santos, given=Luiz Bonino, prefix=da, useprefix=true and Bourne, Philip E. and Bouwman, Jildau and Brookes, Anthony J. and Clark, Tim and Crosas, Mercè and Dillo, Ingrid and Dumon, Olivier and Edmunds, Scott and Evelo, Chris T. and Finkers, Richard and Gonzalez-Beltran, Alejandra and Gray, Alasdair J. G. and Groth, Paul and Goble, Carole and Grethe, Jeffrey S. and Heringa, Jaap and family=Hoen, given=Peter A. C., prefix=’t, useprefix=true and Hooft, Rob and Kuhn, Tobias and Kok, Ruben and Kok, Joost and Lusher, Scott J. and Martone, Maryann E. and Mons, Albert and Packer, Abel L. and Persson, Bengt and Rocca-Serra, Philippe and Roos, Marco and family=Schaik, given=Rene, prefix=van, useprefix=true and Sansone, Susanna-Assunta and Schultes, Erik and Sengstag, Thierry and Slater, Ted and Strawn, George and Swertz, Morris A. and Thompson, Mark and family=Lei, given=Johan, prefix=van der, useprefix=true and family=Mulligen, given=Erik, prefix=van, useprefix=true and Velterop, Jan and Waagmeester, Andra and Wittenburg, Peter and Wolstencroft, Katherine and Zhao, Jun and Mons, Barend},
date = {2016-03-15},
journaltitle = {Scientific Data},
shortjournal = {Sci Data},
@@ -7311,7 +8056,7 @@
langid = {english},
keywords = {FAIR,original publication},
annotation = {Bandiera\_abtest: a Cg\_type: Nature Research Journals Primary\_atype: Comments \& Opinion Subject\_term: Publication characteristics;Research data Subject\_term\_id: publication-characteristics;research-data},
- file = {/home/johannes/Nextcloud/Zotero/Wilkinson et al_2016_The FAIR Guiding Principles for scientific data management and stewardship.pdf;/home/johannes/Zotero/storage/7QCVD3LB/sdata201618.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Wilkinson et al_2016_The FAIR Guiding Principles for scientific data management and stewardship.pdf;/Users/wasmer/Zotero/storage/7QCVD3LB/sdata201618.html}
}
@article{willattAtomdensityRepresentationsMachine2019,
@@ -7330,7 +8075,7 @@
urldate = {2021-05-30},
abstract = {The applications of machine learning techniques to chemistry and materials science become more numerous by the day. The main challenge is to devise representations of atomic systems that are at the same time complete and concise, so as to reduce the number of reference calculations that are needed to predict the properties of different types of materials reliably. This has led to a proliferation of alternative ways to convert an atomic structure into an input for a machine-learning model. We introduce an abstract definition of chemical environments that is based on a smoothed atomic density, using a bra-ket notation to emphasize basis set independence and to highlight the connections with some popular choices of representations for describing atomic systems. The correlations between the spatial distribution of atoms and their chemical identities are computed as inner products between these feature kets, which can be given an explicit representation in terms of the expansion of the atom density on orthogonal basis functions, that is equivalent to the smooth overlap of atomic positions power spectrum, but also in real space, corresponding to n-body correlations of the atom density. This formalism lays the foundations for a more systematic tuning of the behavior of the representations, by introducing operators that represent the correlations between structure, composition, and the target properties. It provides a unifying picture of recent developments in the field and indicates a way forward toward more effective and computationally affordable machine-learning schemes for molecules and materials.},
keywords = {ACSF,alchemical,chemical species scaling problem,descriptor dimred,descriptors,descriptors analysis,dimensionality reduction,MBTR,ML,SOAP},
- file = {/home/johannes/Nextcloud/Zotero/Willatt et al_2019_Atom-density representations for machine learning.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Willatt et al_2019_Atom-density representations for machine learning.pdf}
}
@article{willattFeatureOptimizationAtomistic2018,
@@ -7350,7 +8095,7 @@
abstract = {Machine-learning of atomic-scale properties amounts to extracting correlations between structure, composition and the quantity that one wants to predict. Representing the input structure in a way that best reflects such correlations makes it possible to improve the accuracy of the model for a given amount of reference data. When using a description of the structures that is transparent and well-principled, optimizing the representation might reveal insights into the chemistry of the data set. Here we show how one can generalize the SOAP kernel to introduce a distance-dependent weight that accounts for the multi-scale nature of the interactions, and a description of correlations between chemical species. We show that this improves substantially the performance of ML models of molecular and materials stability, while making it easier to work with complex, multi-component systems and to extend SOAP to coarse-grained intermolecular potentials. The element correlations that give the best performing model show striking similarities with the conventional periodic table of the elements, providing an inspiring example of how machine learning can rediscover, and generalize, intuitive concepts that constitute the foundations of chemistry.},
langid = {english},
keywords = {alchemical,chemical species scaling problem,descriptor dimred,descriptors,dimensionality reduction,ML,SOAP},
- file = {/home/johannes/Nextcloud/Zotero/Willatt et al_2018_Feature optimization for atomistic machine learning yields a data-driven.pdf;/home/johannes/Zotero/storage/ZY2VC9JE/C8CP05921G.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Willatt et al_2018_Feature optimization for atomistic machine learning yields a data-driven.pdf;/Users/wasmer/Zotero/storage/ZY2VC9JE/C8CP05921G.html}
}
@unpublished{winterUnsupervisedLearningGroup2022,
@@ -7359,13 +8104,12 @@
date = {2022-02-15},
eprint = {2202.07559},
eprinttype = {arxiv},
- primaryclass = {cs},
+ eprintclass = {cs},
url = {http://arxiv.org/abs/2202.07559},
urldate = {2022-05-11},
abstract = {Equivariant neural networks, whose hidden features transform according to representations of a group G acting on the data, exhibit training efficiency and an improved generalisation performance. In this work, we extend group invariant and equivariant representation learning to the field of unsupervised deep learning. We propose a general learning strategy based on an encoder-decoder framework in which the latent representation is disentangled in an invariant term and an equivariant group action component. The key idea is that the network learns the group action on the data space and thus is able to solve the reconstruction task from an invariant data representation, hence avoiding the necessity of ad-hoc group-specific implementations. We derive the necessary conditions on the equivariant encoder, and we present a construction valid for any G, both discrete and continuous. We describe explicitly our construction for rotations, translations and permutations. We test the validity and the robustness of our approach in a variety of experiments with diverse data types employing different network architectures.},
- archiveprefix = {arXiv},
keywords = {Computer Science - Machine Learning},
- file = {/home/johannes/Nextcloud/Zotero/Winter et al_2022_Unsupervised Learning of Group Invariant and Equivariant Representations.pdf;/home/johannes/Zotero/storage/5PYE8XM2/2202.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Winter et al_2022_Unsupervised Learning of Group Invariant and Equivariant Representations.pdf;/Users/wasmer/Zotero/storage/5PYE8XM2/2202.html}
}
@online{woodQuantumComplexityTamed2022,
@@ -7378,7 +8122,7 @@
langid = {english},
organization = {{Quanta Magazine}},
keywords = {DeepMind,DFT,DM21,for introductions,ML-DFA,ML-DFT,ML-ESM,molecules,popular science},
- file = {/home/johannes/Zotero/storage/Y7SURWT5/quantum-complexity-tamed-by-machine-learning-20220207.html}
+ file = {/Users/wasmer/Zotero/storage/Y7SURWT5/quantum-complexity-tamed-by-machine-learning-20220207.html}
}
@article{wurgerExploringStructurepropertyRelationships2021,
@@ -7400,7 +8144,7 @@
langid = {english},
keywords = {descriptor comparison,descriptors,dimensionality reduction,KRR,materials discovery,ML,models,sketchmap,SOAP},
annotation = {Bandiera\_abtest: a Cc\_license\_type: cc\_by Cg\_type: Nature Research Journals Primary\_atype: Research Subject\_term: Computational methods;Corrosion;Mathematics and computing;Theoretical chemistry Subject\_term\_id: computational-methods;corrosion;mathematics-and-computing;theoretical-chemistry},
- file = {/home/johannes/Nextcloud/Zotero/Würger et al_2021_Exploring structure-property relationships in magnesium dissolution modulators.pdf;/home/johannes/Zotero/storage/NM6RVQRY/s41529-020-00148-z.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Würger et al_2021_Exploring structure-property relationships in magnesium dissolution modulators.pdf;/Users/wasmer/Zotero/storage/NM6RVQRY/s41529-020-00148-z.html}
}
@article{xieCrystalGraphConvolutional2018,
@@ -7418,7 +8162,7 @@
urldate = {2022-09-27},
abstract = {The use of machine learning methods for accelerating the design of crystalline materials usually requires manually constructed feature vectors or complex transformation of atom coordinates to input the crystal structure, which either constrains the model to certain crystal types or makes it difficult to provide chemical insights. Here, we develop a crystal graph convolutional neural networks framework to directly learn material properties from the connection of atoms in the crystal, providing a universal and interpretable representation of crystalline materials. Our method provides a highly accurate prediction of density functional theory calculated properties for eight different properties of crystals with various structure types and compositions after being trained with 104 data points. Further, our framework is interpretable because one can extract the contributions from local chemical environments to global properties. Using an example of perovskites, we show how this information can be utilized to discover empirical rules for materials design.},
keywords = {_tablet,CGCNN,GCN,GNN,library,solids,with-code},
- file = {/home/johannes/Nextcloud/Zotero/Xie_Grossman_2018_Crystal Graph Convolutional Neural Networks for an Accurate and Interpretable.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Xie_Grossman_2018_Crystal Graph Convolutional Neural Networks for an Accurate and Interpretable.pdf}
}
@unpublished{xieUltrafastInterpretableMachinelearning2021,
@@ -7427,13 +8171,12 @@
date = {2021-10-01},
eprint = {2110.00624},
eprinttype = {arxiv},
- primaryclass = {cond-mat, physics:physics},
+ eprintclass = {cond-mat, physics:physics},
url = {http://arxiv.org/abs/2110.00624},
urldate = {2022-05-09},
abstract = {All-atom dynamics simulations are an indispensable quantitative tool in physics, chemistry, and materials science, but large systems and long simulation times remain challenging due to the trade-off between computational efficiency and predictive accuracy. To address this challenge, we combine effective two- and three-body potentials in a cubic B-spline basis with regularized linear regression to obtain machine-learning potentials that are physically interpretable, sufficiently accurate for applications, as fast as the fastest traditional empirical potentials, and two to four orders of magnitude faster than state-of-the-art machine-learning potentials. For data from empirical potentials, we demonstrate exact retrieval of the potential. For data from density functional theory, the predicted energies, forces, and derived properties, including phonon spectra, elastic constants, and melting points, closely match those of the reference method. The introduced potentials might contribute towards accurate all-atom dynamics simulations of large atomistic systems over long time scales.},
- archiveprefix = {arXiv},
keywords = {descriptors,UFP},
- file = {/home/johannes/Nextcloud/Zotero/Xie et al_2021_Ultra-fast interpretable machine-learning potentials.pdf;/home/johannes/Zotero/storage/8585X9YA/2110.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Xie et al_2021_Ultra-fast interpretable machine-learning potentials.pdf;/Users/wasmer/Zotero/storage/8585X9YA/2110.html}
}
@article{xuSurveyMultiOutputLearning2020,
@@ -7449,7 +8192,7 @@
abstract = {The aim of multi-output learning is to simultaneously predict multiple outputs given an input. It is an important learning problem for decision-making since making decisions in the real world often involves multiple complex factors and criteria. In recent times, an increasing number of research studies have focused on ways to predict multiple outputs at once. Such efforts have transpired in different forms according to the particular multi-output learning problem under study. Classic cases of multi-output learning include multi-label learning, multi-dimensional learning, multi-target regression, and others. From our survey of the topic, we were struck by a lack in studies that generalize the different forms of multi-output learning into a common framework. This article fills that gap with a comprehensive review and analysis of the multi-output learning paradigm. In particular, we characterize the four Vs of multi-output learning, i.e., volume, velocity, variety, and veracity, and the ways in which the four Vs both benefit and bring challenges to multi-output learning by taking inspiration from big data. We analyze the life cycle of output labeling, present the main mathematical definitions of multi-output learning, and examine the field's key challenges and corresponding solutions as found in the literature. Several model evaluation metrics and popular data repositories are also discussed. Last but not least, we highlight some emerging challenges with multi-output learning from the perspective of the four Vs as potential research directions worthy of further studies.},
eventtitle = {{{IEEE Transactions}} on {{Neural Networks}} and {{Learning Systems}}},
keywords = {ML,multi-output learning,multi-target learning,output label representation,structured output prediction,Supervised learning},
- file = {/home/johannes/Nextcloud/Zotero/Xu et al_2020_Survey on Multi-Output Learning.pdf;/home/johannes/Zotero/storage/9TWMMATA/8892612.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Xu et al_2020_Survey on Multi-Output Learning.pdf;/Users/wasmer/Zotero/storage/9TWMMATA/8892612.html}
}
@article{yamadaPredictingMaterialsProperties2019,
@@ -7468,7 +8211,7 @@
urldate = {2021-05-15},
abstract = {There is a growing demand for the use of machine learning (ML) to derive fast-to-evaluate surrogate models of materials properties. In recent years, a broad array of materials property databases have emerged as part of a digital transformation of materials science. However, recent technological advances in ML are not fully exploited because of the insufficient volume and diversity of materials data. An ML framework called “transfer learning†has considerable potential to overcome the problem of limited amounts of materials data. Transfer learning relies on the concept that various property types, such as physical, chemical, electronic, thermodynamic, and mechanical properties, are physically interrelated. For a given target property to be predicted from a limited supply of training data, models of related proxy properties are pretrained using sufficient data; these models capture common features relevant to the target task. Repurposing of such machine-acquired features on the target task yields outstanding prediction performance even with exceedingly small data sets, as if highly experienced human experts can make rational inferences even for considerably less experienced tasks. In this study, to facilitate widespread use of transfer learning, we develop a pretrained model library called XenonPy.MDL. In this first release, the library comprises more than 140\,000 pretrained models for various properties of small molecules, polymers, and inorganic crystalline materials. Along with these pretrained models, we describe some outstanding successes of transfer learning in different scenarios such as building models with only dozens of materials data, increasing the ability of extrapolative prediction through a strategic model transfer, and so on. Remarkably, transfer learning has autonomously identified rather nontrivial transferability across different properties transcending the different disciplines of materials science; for example, our analysis has revealed underlying bridges between small molecules and polymers and between organic and inorganic chemistry.},
keywords = {compositional descriptors,database-based descriptors,descriptors,library,ML,models,notebooks,OFM descriptor,pretrained models,python,pytorch,RDF descriptor,small data,transfer learning,visualization,with-code,XenonPy},
- file = {/home/johannes/Nextcloud/Zotero/Yamada et al_2019_Predicting Materials Properties with Little Data Using Shotgun Transfer Learning.pdf;/home/johannes/Zotero/storage/4F8PQPMD/acscentsci.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Yamada et al_2019_Predicting Materials Properties with Little Data Using Shotgun Transfer Learning.pdf;/Users/wasmer/Zotero/storage/4F8PQPMD/acscentsci.html}
}
@article{yangMachinelearningAcceleratedGeometry2021,
@@ -7487,7 +8230,7 @@
urldate = {2021-06-24},
abstract = {Geometry optimization is an important part of both computational materials and surface science because it is the path to finding ground state atomic structures and reaction pathways. These properties are used in the estimation of thermodynamic and kinetic properties of molecular and crystal structures. This process is slow at the quantum level of theory because it involves an iterative calculation of forces using quantum chemical codes such as density functional theory (DFT), which are computationally expensive and which limit the speed of the optimization algorithms. It would be highly advantageous to accelerate this process because then one could do either the same amount of work in less time or more work in the same time. In this work, we provide a neural network (NN) ensemble based active learning method to accelerate the local geometry optimization for multiple configurations simultaneously. We illustrate the acceleration on several case studies including bare metal surfaces, surfaces with adsorbates, and nudged elastic band for two reactions. In all cases, the accelerated method requires fewer DFT calculations than the standard method. In addition, we provide an Atomic Simulation Environment (ASE)-optimizer Python package to make the usage of the NN ensemble active learning for geometry optimization easier.},
keywords = {ACSF,active learning,BPNN,DFT,GPR,ML,SingleNN,structure relaxation,surrogate model,to_read},
- file = {/home/johannes/Nextcloud/Zotero/false;/home/johannes/Zotero/storage/2L5JFJN8/5.html}
+ file = {/home/johannes/Nextcloud/Zotero/false;/Users/wasmer/Zotero/storage/2L5JFJN8/5.html}
}
@online{zachglickDoesItFeel2021,
@@ -7500,7 +8243,7 @@
langid = {english},
organization = {{@ZachLGlick}},
keywords = {ML},
- file = {/home/johannes/Zotero/storage/WCHKALVA/1384860348730298375.html}
+ file = {/Users/wasmer/Zotero/storage/WCHKALVA/1384860348730298375.html}
}
@article{zahariaAcceleratingMachineLearning2018,
@@ -7509,7 +8252,7 @@
date = {2018},
journaltitle = {IEEE Data Eng. Bull.},
abstract = {MLflow, an open source platform recently launched to streamline the machine learning lifecycle, covers three key challenges: experimentation, reproducibility, and model deployment, using generic APIs that work with any ML library, algorithm and programming language. Machine learning development creates multiple new challenges that are not present in a traditional software development lifecycle. These include keeping track of the myriad inputs to an ML application (e.g., data versions, code and tuning parameters), reproducing results, and production deployment. In this paper, we summarize these challenges from our experience with Databricks customers, and describe MLflow, an open source platform we recently launched to streamline the machine learning lifecycle. MLflow covers three key challenges: experimentation, reproducibility, and model deployment, using generic APIs that work with any ML library, algorithm and programming language. The project has a rapidly growing open source community, with over 50 contributors since its launch in June 2018.},
- file = {/home/johannes/Nextcloud/Zotero/Zaharia et al_2018_Accelerating the Machine Learning Lifecycle with MLflow.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Zaharia et al_2018_Accelerating the Machine Learning Lifecycle with MLflow.pdf}
}
@article{zaverkinFastSampleEfficientInteratomic2021,
@@ -7527,7 +8270,7 @@
url = {https://doi.org/10.1021/acs.jctc.1c00527},
urldate = {2022-01-02},
abstract = {Artificial neural networks (NNs) are one of the most frequently used machine learning approaches to construct interatomic potentials and enable efficient large-scale atomistic simulations with almost ab initio accuracy. However, the simultaneous training of NNs on energies and forces, which are a prerequisite for, e.g., molecular dynamics simulations, can be demanding. In this work, we present an improved NN architecture based on the previous GM-NN model [Zaverkin V.; Kästner, J. J. Chem. Theory Comput. 2020, 16, 5410−5421], which shows an improved prediction accuracy and considerably reduced training times. Moreover, we extend the applicability of Gaussian moment-based interatomic potentials to periodic systems and demonstrate the overall excellent transferability and robustness of the respective models. The fast training by the improved methodology is a prerequisite for training-heavy workflows such as active learning or learning-on-the-fly.},
- file = {/home/johannes/Nextcloud/Zotero/Zaverkin et al_2021_Fast and Sample-Efficient Interatomic Neural Network Potentials for Molecules.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Zaverkin et al_2021_Fast and Sample-Efficient Interatomic Neural Network Potentials for Molecules.pdf}
}
@article{zeledonStructuralInformationFiltered2020,
@@ -7547,7 +8290,7 @@
urldate = {2021-05-15},
abstract = {Machine learning inspired potentials continue to improve the ability for predicting structures of materials. However, many challenges still exist, particularly when calculating structures of disordered systems. These challenges are primarily due to the rapidly increasing dimensionality of the feature-vector space which in most machine-learning algorithms is dependent on the size of the structure. In this article, we present a feature-engineered approach that establishes a set of principles for representing potentials of physical structures (crystals, molecules, and clusters) in a feature space rather than a physically motivated space. Our goal in this work is to define guiding principles that optimize information storage of the physical parameters within the feature representations. In this manner, we focus on keeping the dimensionality of the feature space independent of the number of atoms in the structure. Our Structural Information Filtered Features (SIFF) potential represents structures by utilizing a feature vector of low-correlated descriptors, which correspondingly maximizes information within the descriptor. We present results of our SIFF potential on datasets composed of disordered (carbon and carbon–oxygen) clusters, molecules with C7O2H2 stoichiometry in the GDB9-14B dataset, and crystal structures of the form (AlxGayInz)2O3 as proposed in the NOMAD Kaggle competition. Our potential's performance is at least comparable, sometimes significantly more accurate, and often more efficient than other well-known machine-learning potentials for structure prediction. However, primarily, we offer a different perspective on how researchers should consider opportunities in maximizing information storage for features.},
keywords = {Behler,Behler-Parrinello potential,descriptors,feature engineering,ML,MLP,SIFF},
- file = {/home/johannes/Nextcloud/Zotero/Zeledon et al_2020_The structural information filtered features (SIFF) potential.pdf;/home/johannes/Zotero/storage/DQJJR84B/5.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Zeledon et al_2020_The structural information filtered features (SIFF) potential.pdf;/Users/wasmer/Zotero/storage/DQJJR84B/5.html}
}
@report{zellerCorrelatedElectronsModels2012,
@@ -7564,14 +8307,13 @@
editorbtype = {redactor},
isbn = {9783893367962},
langid = {english},
- keywords = {_tablet,KKR,PGI-1/IAS-1},
- file = {/home/johannes/Nextcloud/Zotero/Zeller_2012_Correlated electrons.pdf;/home/johannes/Zotero/storage/BKBRXSWN/136393.html}
+ keywords = {_tablet,CPA,KKR,PGI-1/IAS-1,VCA},
+ file = {/Users/wasmer/Nextcloud/Zotero/Zeller_2012_Correlated electrons.pdf;/Users/wasmer/Zotero/storage/BKBRXSWN/136393.html}
}
@article{zeniCompactAtomicDescriptors2021,
title = {Compact Atomic Descriptors Enable Accurate Predictions via Linear Models},
- author = {Zeni, Claudio and Rossi, Kevin and Glielmo, Aldo and de Gironcoli, Stefano},
- options = {useprefix=true},
+ author = {Zeni, Claudio and Rossi, Kevin and Glielmo, Aldo and family=Gironcoli, given=Stefano, prefix=de, useprefix=true},
date = {2021-06-14},
journaltitle = {The Journal of Chemical Physics},
shortjournal = {J. Chem. Phys.},
@@ -7585,7 +8327,7 @@
urldate = {2022-05-11},
abstract = {We probe the accuracy of linear ridge regression employing a three-body local density representation derived from the atomic cluster expansion. We benchmark the accuracy of this framework in the prediction of formation energies and atomic forces in molecules and solids. We find that such a simple regression framework performs on par with state-of-the-art machine learning methods which are, in most cases, more complex and more computationally demanding. Subsequently, we look for ways to sparsify the descriptor and further improve the computational efficiency of the method. To this aim, we use both principal component analysis and least absolute shrinkage operator regression for energy fitting on six single-element datasets. Both methods highlight the possibility of constructing a descriptor that is four times smaller than the original with a similar or even improved accuracy. Furthermore, we find that the reduced descriptors share a sizable fraction of their features across the six independent datasets, hinting at the possibility of designing material-agnostic, optimally compressed, and accurate descriptors.},
keywords = {_tablet,ACE,descriptor dimred,descriptors,dimensionality reduction},
- file = {/home/johannes/Nextcloud/Zotero/Zeni et al_2021_Compact atomic descriptors enable accurate predictions via linear models.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Zeni et al_2021_Compact atomic descriptors enable accurate predictions via linear models.pdf}
}
@article{zepeda-nunezDeepDensityCircumventing2021,
@@ -7604,7 +8346,7 @@
abstract = {The recently developed Deep Potential [Phys. Rev. Lett. 120 (2018) 143001 [27]] is a powerful method to represent general inter-atomic potentials using deep neural networks. The success of Deep Potential rests on the proper treatment of locality and symmetry properties of each component of the network. In this paper, we leverage its network structure to effectively represent the mapping from the atomic configuration to the electron density in Kohn-Sham density function theory (KS-DFT). By directly targeting at the self-consistent electron density, we demonstrate that the adapted network architecture, called the Deep Density, can effectively represent the self-consistent electron density as the linear combination of contributions from many local clusters. The network is constructed to satisfy the translation, rotation, and permutation symmetries, and is designed to be transferable to different system sizes. We demonstrate that using a relatively small number of training snapshots, with each snapshot containing a modest amount of data-points, Deep Density achieves excellent performance for one-dimensional insulating and metallic systems, as well as systems with mixed insulating and metallic characters. We also demonstrate its performance for real three-dimensional systems, including small organic molecules, as well as extended systems such as water (up to 512 molecules) and aluminum (up to 256 atoms).},
langid = {english},
keywords = {_tablet,DFT,ML,ML-DFT,ML-ESM,prediction of electron density},
- file = {/home/johannes/Nextcloud/Zotero/Zepeda-Núñez et al_2021_Deep Density.pdf;/home/johannes/Zotero/storage/TJJ4NCEI/S0021999121004186.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Zepeda-Núñez et al_2021_Deep Density.pdf;/Users/wasmer/Zotero/storage/TJJ4NCEI/S0021999121004186.html}
}
@article{zhangDeepPotentialMolecular2018,
@@ -7617,7 +8359,27 @@
volume = {120},
number = {14},
doi = {10.1103/PhysRevLett.120.143001},
- file = {/home/johannes/Nextcloud/Zotero/Zhang_2018_Deep Potential Molecular Dynamics.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Zhang_2018_Deep Potential Molecular Dynamics.pdf}
+}
+
+@article{zhangHighthroughputDesignMagnetic2021,
+ title = {High-Throughput Design of Magnetic Materials},
+ author = {Zhang, Hongbin},
+ date = {2021-09},
+ journaltitle = {Electronic Structure},
+ shortjournal = {Electron. Struct.},
+ volume = {3},
+ number = {3},
+ pages = {033001},
+ publisher = {{IOP Publishing}},
+ issn = {2516-1075},
+ doi = {10.1088/2516-1075/abbb25},
+ url = {https://dx.doi.org/10.1088/2516-1075/abbb25},
+ urldate = {2023-01-25},
+ abstract = {Materials design based on density functional theory (DFT) calculations is an emergent field of great potential to accelerate the development and employment of novel materials. Magnetic materials play an essential role in green energy applications as they provide efficient ways of harvesting, converting, and utilizing energy. In this review, after a brief introduction to the major functionalities of magnetic materials, we demonstrated how the fundamental properties can be tackled via high-throughput DFT calculations, with a particular focus on the current challenges and feasible solutions. Successful case studies are summarized on several classes of magnetic materials, followed by bird-view perspectives.},
+ langid = {english},
+ keywords = {/unread,DFT,electronic structure,Hall QAHE,Hall QSHE,Heisenberg model,High-throughput,magnetic materials,magnetism,magnetocaloric,ML,pgi-1 seminar,review,review-of-magnet,Spintronics,topological insulator,topology and magnetism},
+ file = {/Users/wasmer/Nextcloud/Zotero/Zhang_2021_High-throughput design of magnetic materials.pdf}
}
@article{zhangMachineLearningMathbbZ2017,
@@ -7635,7 +8397,7 @@
urldate = {2021-05-21},
abstract = {After decades of progress and effort, obtaining a phase diagram for a strongly correlated topological system still remains a challenge. Although in principle one could turn to Wilson loops and long-range entanglement, evaluating these nonlocal observables at many points in phase space can be prohibitively costly. With growing excitement over topological quantum computation comes the need for an efficient approach for obtaining topological phase diagrams. Here we turn to machine learning using quantum loop topography (QLT), a notion we have recently introduced. Specifically, we propose a construction of QLT that is sensitive to quasiparticle statistics. We then use mutual statistics between the spinons and visons to detect a Z2 quantum spin liquid in a multiparameter phase space. We successfully obtain the quantum phase boundary between the topological and trivial phases using a simple feed-forward neural network. Furthermore, we demonstrate advantages of our approach for the evaluation of phase diagrams relating to speed and storage. Such statistics-based machine learning of topological phases opens new efficient routes to studying topological phase diagrams in strongly correlated systems.},
keywords = {ANN,ML,quantum spin liquid,topological phase,topological phase transition},
- file = {/home/johannes/Nextcloud/Zotero/Zhang et al_2017_Machine learning $ -mathbb Z _ 2 $ quantum spin liquids with quasiparticle.pdf;/home/johannes/Zotero/storage/39L63FS4/PhysRevB.96.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Zhang et al_2017_Machine learning $ -mathbb Z _ 2 $ quantum spin liquids with quasiparticle.pdf;/Users/wasmer/Zotero/storage/39L63FS4/PhysRevB.96.html}
}
@article{zhangMachineLearningTopological2018,
@@ -7653,7 +8415,24 @@
urldate = {2021-05-21},
abstract = {In this Letter we supervisedly train neural networks to distinguish different topological phases in the context of topological band insulators. After training with Hamiltonians of one-dimensional insulators with chiral symmetry, the neural network can predict their topological winding numbers with nearly 100\% accuracy, even for Hamiltonians with larger winding numbers that are not included in the training data. These results show a remarkable success that the neural network can capture the global and nonlinear topological features of quantum phases from local inputs. By opening up the neural network, we confirm that the network does learn the discrete version of the winding number formula. We also make a couple of remarks regarding the role of the symmetry and the opposite effect of regularization techniques when applying machine learning to physical systems.},
keywords = {ANN,ML,regularization,symmetry,topological phase},
- file = {/home/johannes/Nextcloud/Zotero/Zhang et al_2018_Machine Learning Topological Invariants with Neural Networks.pdf;/home/johannes/Zotero/storage/XCPMLTVF/PhysRevLett.120.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Zhang et al_2018_Machine Learning Topological Invariants with Neural Networks.pdf;/Users/wasmer/Zotero/storage/XCPMLTVF/PhysRevLett.120.html}
+}
+
+@online{zhangPushingLimitsAtomistic2019,
+ title = {Pushing the Limits of Atomistic Simulations towards Ultra-High Temperature: A Machine-Learning Force Field for {{ZrB2}}},
+ shorttitle = {Pushing the Limits of Atomistic Simulations towards Ultra-High Temperature},
+ author = {Zhang, Yanhui and Lunghi, Alessandro and Sanvito, Stefano},
+ date = {2019-11-08},
+ number = {arXiv:1911.03307},
+ eprint = {arXiv:1911.03307},
+ eprinttype = {arxiv},
+ doi = {10.48550/arXiv.1911.03307},
+ url = {http://arxiv.org/abs/1911.03307},
+ urldate = {2023-02-23},
+ abstract = {Determining thermal and physical quantities across a broad temperature domain, especially up to the ultra-high temperature region, is a formidable theoretical and experimental challenge. At the same time it is essential for understanding the performance of ultra-high temperature ceramic (UHTC) materials. Here we present the development of a machine-learning force field for ZrB2, one of the primary members of the UHTC family with a complex bonding structure. The force field exhibits chemistry accuracy for both energies and forces and can reproduce structural, elastic and phonon properties, including thermal expansion and thermal transport. A thorough comparison with available empirical potentials shows that our force field outperforms the competitors. Most importantly, its effectiveness is extended from room temperature to the ultra-high temperature region (up to \textasciitilde{} 2,500 K), where measurements are very difficult, costly and some time impossible. Our work demonstrates that machine-learning force fields can be used for simulations of materials in a harsh environment, where no experimental tools are available, but crucial for a number of engineering applications, such as in aerospace, aviation and nuclear.},
+ pubstate = {preprint},
+ keywords = {_tablet,/unread,Condensed Matter - Materials Science,Quantum Physics},
+ file = {/Users/wasmer/Nextcloud/Zotero/Zhang et al_2019_Pushing the limits of atomistic simulations towards ultra-high temperature.pdf;/Users/wasmer/Zotero/storage/IIZUJ8Y9/1911.html}
}
@article{zhangQuantumLoopTopography2017,
@@ -7671,7 +8450,7 @@
urldate = {2021-05-21},
abstract = {Despite rapidly growing interest in harnessing machine learning in the study of quantum many-body systems, training neural networks to identify quantum phases is a nontrivial challenge. The key challenge is in efficiently extracting essential information from the many-body Hamiltonian or wave function and turning the information into an image that can be fed into a neural network. When targeting topological phases, this task becomes particularly challenging as topological phases are defined in terms of nonlocal properties. Here, we introduce quantum loop topography (QLT): a procedure of constructing a multidimensional image from the “sample†Hamiltonian or wave function by evaluating two-point operators that form loops at independent Monte Carlo steps. The loop configuration is guided by the characteristic response for defining the phase, which is Hall conductivity for the cases at hand. Feeding QLT to a fully connected neural network with a single hidden layer, we demonstrate that the architecture can be effectively trained to distinguish the Chern insulator and the fractional Chern insulator from trivial insulators with high fidelity. In addition to establishing the first case of obtaining a phase diagram with a topological quantum phase transition with machine learning, the perspective of bridging traditional condensed matter theory with machine learning will be broadly valuable.},
keywords = {ANN,classification,ML,QLT,QMC,topological phase transition},
- file = {/home/johannes/Nextcloud/Zotero/Zhang_Kim_2017_Quantum Loop Topography for Machine Learning.pdf;/home/johannes/Zotero/storage/M9VFL53W/PhysRevLett.118.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Zhang_Kim_2017_Quantum Loop Topography for Machine Learning.pdf;/Users/wasmer/Zotero/storage/M9VFL53W/PhysRevLett.118.html}
}
@article{zhangStrategyApplyMachine2018,
@@ -7691,7 +8470,7 @@
issue = {1},
langid = {english},
keywords = {KRR,ML,models,rec-by-ruess,small data},
- file = {/home/johannes/Nextcloud/Zotero/Zhang_Ling_2018_A strategy to apply machine learning to small datasets in materials science.pdf;/home/johannes/Zotero/storage/PEGZREYC/s41524-018-0081-z.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Zhang_Ling_2018_A strategy to apply machine learning to small datasets in materials science.pdf;/Users/wasmer/Zotero/storage/PEGZREYC/s41524-018-0081-z.html}
}
@article{zhaoQuantumOscillationsIrondoped2019,
@@ -7708,7 +8487,7 @@
url = {https://link.aps.org/doi/10.1103/PhysRevB.99.165133},
urldate = {2022-04-28},
abstract = {We investigated the magnetotransport properties of Fe-doped topological insulator Sb1.96Fe0.04Te3 single crystals. With doping, the band structure changes significantly and multiple Fermi pockets become evident in the Shubnikov–de Haas oscillations, in contrast to the single frequency detected for pure Sb2Te3. Using complementary density functional theory calculations, we identify an additional bulk hole pocket introduced at the Γ point which originates from the chemical distortion associated with the Fe dopant. Experimentally, both doped and undoped samples are hole-carrier dominated; however, Fe doping also reduces the carrier density and mobility. The angle dependent quantum oscillations of Sb1.96Fe0.04Te3 were analyzed to characterize the complex Fermi surface and isolate the dimensionality of each SdH feature. Among those components, we found two oscillations frequencies, which related to two Fermi pockets are highly angle dependent. Moreover, the fermiology changes via Fe doping and may also provide a different Berry phase, as demonstrated by the Landau fan diagram, thus indicating a rich complexity in the underlying electronic structure.},
- file = {/home/johannes/Nextcloud/Zotero/Zhao et al_2019_Quantum oscillations in iron-doped single crystals of the topological insulator.pdf;/home/johannes/Zotero/storage/GGTED6FM/Zhao et al. - 2019 - Quantum oscillations in iron-doped single crystals.pdf;/home/johannes/Zotero/storage/8D5JL2DQ/PhysRevB.99.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Zhao et al_2019_Quantum oscillations in iron-doped single crystals of the topological insulator.pdf;/Users/wasmer/Zotero/storage/GGTED6FM/Zhao et al. - 2019 - Quantum oscillations in iron-doped single crystals.pdf;/Users/wasmer/Zotero/storage/8D5JL2DQ/PhysRevB.99.html}
}
@thesis{zimmermannInitioDescriptionTransverse2014,
@@ -7723,7 +8502,7 @@
isbn = {9783893369850},
langid = {english},
keywords = {_tablet,Boltzmann theory,Hall AHE,Hall effect,Hall SHE,juKKR,KKR,PGI-1/IAS-1,thesis,transport properties},
- file = {/home/johannes/Nextcloud/Zotero/Zimmermann_2014_Ab initio description of transverse transport due to impurity scattering in.pdf;/home/johannes/Zotero/storage/QL7I6VYG/171881.html}
+ file = {/Users/wasmer/Nextcloud/Zotero/Zimmermann_2014_Ab initio description of transverse transport due to impurity scattering in.pdf;/Users/wasmer/Zotero/storage/QL7I6VYG/171881.html}
}
@article{zungerUnderstandingDopingQuantum2021,
@@ -7741,7 +8520,5 @@
url = {https://doi.org/10.1021/acs.chemrev.0c00608},
urldate = {2021-08-23},
abstract = {Doping mobile carriers into ordinary semiconductors such as Si, GaAs, and ZnO was the enabling step in the electronic and optoelectronic revolutions. The recent emergence of a class of “quantum materialsâ€, where uniquely quantum interactions between the components produce specific behaviors such as topological insulation, unusual magnetism, superconductivity, spin–orbit-induced and magnetically induced spin splitting, polaron formation, and transparency of electrical conductors, pointed attention to a range of doping-related phenomena associated with chemical classes that differ from the traditional semiconductors. These include wide-gap oxides, compounds containing open-shell d electrons, and compounds made of heavy elements yet having significant band gaps. The atomistic electronic structure theory of doping that has been developed over the past two decades in the subfield of semiconductor physics has recently been extended and applied to quantum materials. The present review focuses on explaining the main concepts needed for a basic understanding of the doping phenomenology and indeed peculiarities in quantum materials from the perspective of condensed matter theory, with the hope of forging bridges to the chemists that have enabled the synthesis of some of the most interesting compounds in this field.},
- file = {/home/johannes/Nextcloud/Zotero/Zunger_Malyi_2021_Understanding Doping of Quantum Materials.pdf}
+ file = {/Users/wasmer/Nextcloud/Zotero/Zunger_Malyi_2021_Understanding Doping of Quantum Materials.pdf}
}
-
-@preamble{ "\ifdefined\DeclarePrefChars\DeclarePrefChars{'’-}\else\fi " }
--
GitLab