Local invertibility and sensitivity of atomic structure-feature mappings [version 1; peer review: 2 approved]

Background: The increasingly common applications of machine-learning schemes to atomic-scale simulations have triggered efforts to better understand the mathematical properties of the mapping between the Cartesian coordinates of the atoms and the variety of representations that can be used to conver...

Descripción completa

Guardado en:
Detalles Bibliográficos
Autores principales: Sergey N. Pozdnyakov, Liwei Zhang, Christoph Ortner, Gábor Csányi, Michele Ceriotti
Formato: article
Lenguaje:EN
Publicado: F1000 Research Limited 2021
Materias:
Q
H
Acceso en línea:https://doaj.org/article/a3898f2acb6942dd87685f4e26aaf998
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
id oai:doaj.org-article:a3898f2acb6942dd87685f4e26aaf998
record_format dspace
spelling oai:doaj.org-article:a3898f2acb6942dd87685f4e26aaf9982021-11-22T12:41:10ZLocal invertibility and sensitivity of atomic structure-feature mappings [version 1; peer review: 2 approved]2732-512110.12688/openreseurope.14156.1https://doaj.org/article/a3898f2acb6942dd87685f4e26aaf9982021-10-01T00:00:00Zhttps://open-research-europe.ec.europa.eu/articles/1-126/v1https://doaj.org/toc/2732-5121Background: The increasingly common applications of machine-learning schemes to atomic-scale simulations have triggered efforts to better understand the mathematical properties of the mapping between the Cartesian coordinates of the atoms and the variety of representations that can be used to convert them into a finite set of symmetric descriptors or features. Methods: Here, we analyze the sensitivity of the mapping to atomic displacements, using a singular value decomposition of the Jacobian of the transformation to quantify the sensitivity for different configurations, choice of representations and implementation details.  Results: We show that the combination of symmetry and smoothness leads to mappings that have singular points at which the Jacobian has one or more null singular values (besides those corresponding to infinitesimal translations and rotations). This is in fact desirable, because it enforces physical symmetry constraints on the values predicted by regression models constructed using such representations. However, besides these symmetry-induced singularities, there are also spurious singular points, that we find to be linked to the incompleteness of the mapping, i.e. the fact that, for certain classes of representations, structurally distinct configurations are not guaranteed to be mapped onto different feature vectors. Additional singularities can be introduced by a too aggressive truncation of the infinite basis set that is used to discretize the representations. Conclusions: These results exemplify the subtle issues that arise when constructing symmetric representations of atomic structures, and provide conceptual and numerical tools to identify and investigate them in both benchmark and realistic applications.Sergey N. PozdnyakovLiwei ZhangChristoph OrtnerGábor CsányiMichele CeriottiF1000 Research LimitedarticleScienceQSocial SciencesHENOpen Research Europe, Vol 1 (2021)
institution DOAJ
collection DOAJ
language EN
topic Science
Q
Social Sciences
H
spellingShingle Science
Q
Social Sciences
H
Sergey N. Pozdnyakov
Liwei Zhang
Christoph Ortner
Gábor Csányi
Michele Ceriotti
Local invertibility and sensitivity of atomic structure-feature mappings [version 1; peer review: 2 approved]
description Background: The increasingly common applications of machine-learning schemes to atomic-scale simulations have triggered efforts to better understand the mathematical properties of the mapping between the Cartesian coordinates of the atoms and the variety of representations that can be used to convert them into a finite set of symmetric descriptors or features. Methods: Here, we analyze the sensitivity of the mapping to atomic displacements, using a singular value decomposition of the Jacobian of the transformation to quantify the sensitivity for different configurations, choice of representations and implementation details.  Results: We show that the combination of symmetry and smoothness leads to mappings that have singular points at which the Jacobian has one or more null singular values (besides those corresponding to infinitesimal translations and rotations). This is in fact desirable, because it enforces physical symmetry constraints on the values predicted by regression models constructed using such representations. However, besides these symmetry-induced singularities, there are also spurious singular points, that we find to be linked to the incompleteness of the mapping, i.e. the fact that, for certain classes of representations, structurally distinct configurations are not guaranteed to be mapped onto different feature vectors. Additional singularities can be introduced by a too aggressive truncation of the infinite basis set that is used to discretize the representations. Conclusions: These results exemplify the subtle issues that arise when constructing symmetric representations of atomic structures, and provide conceptual and numerical tools to identify and investigate them in both benchmark and realistic applications.
format article
author Sergey N. Pozdnyakov
Liwei Zhang
Christoph Ortner
Gábor Csányi
Michele Ceriotti
author_facet Sergey N. Pozdnyakov
Liwei Zhang
Christoph Ortner
Gábor Csányi
Michele Ceriotti
author_sort Sergey N. Pozdnyakov
title Local invertibility and sensitivity of atomic structure-feature mappings [version 1; peer review: 2 approved]
title_short Local invertibility and sensitivity of atomic structure-feature mappings [version 1; peer review: 2 approved]
title_full Local invertibility and sensitivity of atomic structure-feature mappings [version 1; peer review: 2 approved]
title_fullStr Local invertibility and sensitivity of atomic structure-feature mappings [version 1; peer review: 2 approved]
title_full_unstemmed Local invertibility and sensitivity of atomic structure-feature mappings [version 1; peer review: 2 approved]
title_sort local invertibility and sensitivity of atomic structure-feature mappings [version 1; peer review: 2 approved]
publisher F1000 Research Limited
publishDate 2021
url https://doaj.org/article/a3898f2acb6942dd87685f4e26aaf998
work_keys_str_mv AT sergeynpozdnyakov localinvertibilityandsensitivityofatomicstructurefeaturemappingsversion1peerreview2approved
AT liweizhang localinvertibilityandsensitivityofatomicstructurefeaturemappingsversion1peerreview2approved
AT christophortner localinvertibilityandsensitivityofatomicstructurefeaturemappingsversion1peerreview2approved
AT gaborcsanyi localinvertibilityandsensitivityofatomicstructurefeaturemappingsversion1peerreview2approved
AT micheleceriotti localinvertibilityandsensitivityofatomicstructurefeaturemappingsversion1peerreview2approved
_version_ 1718417629056598016