{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,2,2]],"date-time":"2026-02-02T21:04:43Z","timestamp":1770066283537,"version":"3.49.0"},"reference-count":64,"publisher":"IOP Publishing","issue":"2","license":[{"start":{"date-parts":[[2021,4,14]],"date-time":"2021-04-14T00:00:00Z","timestamp":1618358400000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2021,4,14]],"date-time":"2021-04-14T00:00:00Z","timestamp":1618358400000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/iopscience.iop.org\/info\/page\/text-and-data-mining"}],"funder":[{"DOI":"10.13039\/501100001711","name":"Schweizerischer Nationalfonds zur F\u00f6rderung der Wissenschaftlichen Forschung","doi-asserted-by":"crossref","award":["Project No. 200021-182057"],"award-info":[{"award-number":["Project No. 200021-182057"]}],"id":[{"id":"10.13039\/501100001711","id-type":"DOI","asserted-by":"crossref"}]},{"DOI":"10.13039\/100010661","name":"Horizon 2020 Framework Programme","doi-asserted-by":"crossref","award":["MaX Center of Excellence"],"award-info":[{"award-number":["MaX Center of Excellence"]}],"id":[{"id":"10.13039\/100010661","id-type":"DOI","asserted-by":"crossref"}]}],"content-domain":{"domain":["iopscience.iop.org"],"crossmark-restriction":false},"short-container-title":["Mach. Learn.: Sci. Technol."],"published-print":{"date-parts":[[2021,6,1]]},"abstract":"<jats:title>Abstract<\/jats:title>\n               <jats:p>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 <jats:italic>n<\/jats:italic>-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\u2013Parrinello 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.<\/jats:p>","DOI":"10.1088\/2632-2153\/abdaf7","type":"journal-article","created":{"date-parts":[[2021,1,14]],"date-time":"2021-01-14T07:50:34Z","timestamp":1610610634000},"page":"025028","update-policy":"https:\/\/doi.org\/10.1088\/crossmark-policy","source":"Crossref","is-referenced-by-count":32,"title":["The role of feature space in atomistic learning"],"prefix":"10.1088","volume":"2","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-8076-215X","authenticated-orcid":false,"given":"Alexander","family":"Goscinski","sequence":"first","affiliation":[]},{"ORCID":"https:\/\/orcid.org\/0000-0003-4824-6512","authenticated-orcid":false,"given":"Guillaume","family":"Fraux","sequence":"additional","affiliation":[]},{"given":"Giulio","family":"Imbalzano","sequence":"additional","affiliation":[]},{"ORCID":"https:\/\/orcid.org\/0000-0003-2571-2832","authenticated-orcid":false,"given":"Michele","family":"Ceriotti","sequence":"additional","affiliation":[]}],"member":"266","published-online":{"date-parts":[[2021,4,14]]},"reference":[{"key":"mlstabdaf7bib1","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevLett.98.146401","volume":"98","author":"Behler","year":"2007","journal-title":"Phys. Rev. Lett."},{"key":"mlstabdaf7bib2","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevLett.104.136403","volume":"104","author":"Bart\u00f3k","year":"2010","journal-title":"Phys. Rev. Lett."},{"key":"mlstabdaf7bib3","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevLett.108.058301","volume":"108","author":"Rupp","year":"2012","journal-title":"Phys. Rev. Lett."},{"key":"mlstabdaf7bib4","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevB.87.184115","volume":"87","author":"Bart\u00f3k","year":"2013","journal-title":"Phys. Rev. B"},{"key":"mlstabdaf7bib5","doi-asserted-by":"publisher","DOI":"10.1039\/C6CP00415F","volume":"18","author":"De","year":"2016","journal-title":"Phys. Chem. Chem. Phys."},{"key":"mlstabdaf7bib6","first-page":"6541","volume":"2017-December","author":"Eickenberg","year":"2017","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"mlstabdaf7bib7","volume":"13754","author":"Huo","year":"2017"},{"key":"mlstabdaf7bib8","doi-asserted-by":"publisher","first-page":"5255","DOI":"10.1021\/acs.jctc.7b00577","volume":"13","author":"Faber","year":"2017","journal-title":"J. Chem. Theory Comput."},{"key":"mlstabdaf7bib9","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1038\/s41467-018-06169-2","volume":"9","author":"Chmiela","year":"2018","journal-title":"Nat. Commun."},{"key":"mlstabdaf7bib10","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevLett.120.143001","volume":"120","author":"Zhang","year":"2018","journal-title":"Phys. Rev. Lett."},{"key":"mlstabdaf7bib11","doi-asserted-by":"publisher","DOI":"10.1063\/1.5090481","volume":"150","author":"Willatt","year":"2019","journal-title":"J. Chem. Phys."},{"key":"mlstabdaf7bib12","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevB.99.014104","volume":"99","author":"Drautz","year":"2019","journal-title":"Phys. Rev. B"},{"key":"mlstabdaf7bib13","doi-asserted-by":"publisher","DOI":"10.1063\/1.5126701","volume":"152","author":"Christensen","year":"2020","journal-title":"J. Chem. Phys."},{"key":"mlstabdaf7bib14","volume":"1","author":"van der Oord","year":"2020","journal-title":"Machine Learning: Science and Technology"},{"key":"mlstabdaf7bib15","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevLett.114.105503","volume":"114","author":"Ghiringhelli","year":"2015","journal-title":"Phys. Rev. Lett."},{"key":"mlstabdaf7bib16","doi-asserted-by":"publisher","DOI":"10.1063\/1.4940026","volume":"144","author":"Zhu","year":"2016","journal-title":"J. Chem. Phys."},{"key":"mlstabdaf7bib17","doi-asserted-by":"publisher","DOI":"10.1063\/1.4818005","volume":"139","author":"Gallet","year":"2013","journal-title":"J. Chem. Phys."},{"key":"mlstabdaf7bib18","doi-asserted-by":"publisher","first-page":"1084","DOI":"10.1002\/qua.24912","volume":"115","author":"von Lilienfeld","year":"2015","journal-title":"Int. J. Quantum Chem."},{"key":"mlstabdaf7bib19","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevLett.125.166001","volume":"125","author":"Pozdnyakov","year":"2020","journal-title":"Phys. Rev. Lett."},{"key":"mlstabdaf7bib20","doi-asserted-by":"publisher","DOI":"10.1039\/C8CP05921G","volume":"20","author":"Willatt","year":"2018","journal-title":"Phys. Chem. Chem. Phys."},{"key":"mlstabdaf7bib21","doi-asserted-by":"publisher","first-page":"731","DOI":"10.1021\/acs.jpca.9b08723","volume":"124","author":"Zuo","year":"2020","journal-title":"J. Phys. Chem. A"},{"key":"mlstabdaf7bib22","doi-asserted-by":"publisher","DOI":"10.1088\/2632-2153\/aba9ef","volume":"1","author":"Helfrecht","year":"2020","journal-title":"Mach. Learn.: Sci. Technol."},{"key":"mlstabdaf7bib23","doi-asserted-by":"publisher","DOI":"10.1063\/5.0016005","volume":"153","author":"Onat","year":"2020","journal-title":"J. Chem. Phys."},{"key":"mlstabdaf7bib24","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevLett.109.059801","volume":"109","author":"Moussa","year":"2012","journal-title":"Phys. Rev. Lett."},{"key":"mlstabdaf7bib25","doi-asserted-by":"publisher","DOI":"10.1063\/1.4828704","volume":"139","author":"Sadeghi","year":"2013","journal-title":"J. Chem. Phys."},{"key":"mlstabdaf7bib26","doi-asserted-by":"publisher","DOI":"10.1088\/2632-2153\/abb212","author":"Parsaeifard","year":"2020","journal-title":"Mach. Learn.: Sci. Technol."},{"key":"mlstabdaf7bib27","first-page":"1415","volume":"3","author":"Torkkola","year":"2003","journal-title":"J. Mach. Learn. Res."},{"key":"mlstabdaf7bib28","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1007\/BF02289451","volume":"31","author":"Sch\u00f6nemann","year":"1966","journal-title":"Psychometrika"},{"key":"mlstabdaf7bib29","doi-asserted-by":"publisher","first-page":"2323","DOI":"10.1126\/science.290.5500.2323","volume":"290","author":"Roweis","year":"2000","journal-title":"Science"},{"key":"mlstabdaf7bib30","doi-asserted-by":"publisher","first-page":"283","DOI":"10.5555\/3008751.3008793","article-title":"The kernel trick for distances","author":"Sch\u00f6lkopf","year":"2001"},{"key":"mlstabdaf7bib31","first-page":"pp 220","author":"Haasdonk","year":"2004"},{"key":"mlstabdaf7bib32","first-page":"pp 583","author":"Sch\u00f6lkopf","year":"1997"},{"key":"mlstabdaf7bib33","doi-asserted-by":"publisher","DOI":"10.1088\/0953-8984\/23\/5\/053201","volume":"23","author":"Pickard","year":"2011","journal-title":"J. Phys. Condens. Matter"},{"key":"mlstabdaf7bib34","article-title":"AIRSS data for carbon at 10 gpa and the C+N+H+O system at 1 gpa","author":"Pickard","year":"2020"},{"key":"mlstabdaf7bib35","doi-asserted-by":"publisher","first-page":"316","DOI":"10.1016\/j.jcp.2014.12.018","volume":"285","author":"Thompson","year":"2015","journal-title":"J. Comput. Phys."},{"key":"mlstabdaf7bib36","article-title":"librascal\u2014A scalable and versatile library to generate representations for atomic-scale learning","author":"Musil"},{"key":"mlstabdaf7bib37","doi-asserted-by":"publisher","DOI":"10.1063\/1.3553717","volume":"134","author":"Behler","year":"2011a","journal-title":"J. Chem. Phys."},{"key":"mlstabdaf7bib38","doi-asserted-by":"crossref","first-page":"3075","DOI":"10.1021\/acs.jctc.8b01092","volume":"15","author":"Singraber","year":"2019","journal-title":"J. Chem. Theory Comput."},{"key":"mlstabdaf7bib39","doi-asserted-by":"publisher","DOI":"10.1039\/c1cp21668f","volume":"13","author":"Behler","year":"2011b","journal-title":"Phys. Chem. Chem. Phys. PCCP"},{"key":"mlstabdaf7bib40","doi-asserted-by":"publisher","DOI":"10.1063\/1.5024611","volume":"148","author":"Imbalzano","year":"2018","journal-title":"J. Chem. Phys."},{"key":"mlstabdaf7bib41","doi-asserted-by":"publisher","DOI":"10.1126\/sciadv.1701816","volume":"3","author":"Bart\u00f3k","year":"2017","journal-title":"Sci. Adv."},{"key":"mlstabdaf7bib42","doi-asserted-by":"publisher","first-page":"4501","DOI":"10.1038\/s41467-018-06972-x","volume":"9","author":"Paruzzo","year":"2018","journal-title":"Nat. Commun."},{"key":"mlstabdaf7bib43","doi-asserted-by":"publisher","DOI":"10.1063\/1.4712397","volume":"136","author":"Jose","year":"2012","journal-title":"J. Chem. Phys."},{"key":"mlstabdaf7bib44","doi-asserted-by":"publisher","first-page":"1032","DOI":"10.1002\/qua.24890","volume":"115","author":"Behler","year":"2015","journal-title":"Int. J. Quantum Chem."},{"key":"mlstabdaf7bib45","doi-asserted-by":"publisher","first-page":"697","DOI":"10.1073\/pnas.0803205106","volume":"106","author":"Mahoney","year":"2009","journal-title":"Proc. Natl Acad. Sci."},{"key":"mlstabdaf7bib46","article-title":"N2P2","author":"Singraber"},{"key":"mlstabdaf7bib47","doi-asserted-by":"crossref","first-page":"1153","DOI":"10.1137\/15M1054183","volume":"14","author":"Shapeev","year":"2016","journal-title":"Multiscale Model. Simul."},{"key":"mlstabdaf7bib48","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevB.97.184307","volume":"97","author":"Glielmo","year":"2018","journal-title":"Phys. Rev. B"},{"key":"mlstabdaf7bib49","doi-asserted-by":"publisher","DOI":"10.1063\/5.0009491","volume":"152","author":"Jinnouchi","year":"2020","journal-title":"J. Chem. Phys."},{"key":"mlstabdaf7bib50","doi-asserted-by":"publisher","DOI":"10.1063\/5.0021116","volume":"153","author":"Nigam","year":"2020","journal-title":"J. Chem. Phys."},{"key":"mlstabdaf7bib51","doi-asserted-by":"publisher","first-page":"784","DOI":"10.1137\/1118101","volume":"18","author":"Vallender","year":"1974","journal-title":"Theory Probab. Its Appl."},{"key":"mlstabdaf7bib52","volume":"1","author":"Cohen","year":"1997"},{"key":"mlstabdaf7bib53","doi-asserted-by":"publisher","first-page":"732","DOI":"10.5555\/1625275.1625393","author":"Cuturi","year":"2007","journal-title":"Int. Jt. Conf. Artif. Intell. IJCAI"},{"key":"mlstabdaf7bib54","doi-asserted-by":"publisher","first-page":"03LT01","DOI":"10.1088\/2632-2153\/aba048","volume":"1","author":"\u00c7aylak","year":"2020","journal-title":"Mach. Learn.: Sci. Technol."},{"key":"mlstabdaf7bib55","doi-asserted-by":"publisher","DOI":"10.1063\/1.4966192","volume":"145","author":"Behler","year":"2016","journal-title":"J. Chem. Phys."},{"key":"mlstabdaf7bib56","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1038\/ncomms15679","volume":"8","author":"Isayev","year":"2017","journal-title":"Nat. Commun."},{"key":"mlstabdaf7bib57","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1038\/s41524-019-0239-3","volume":"5","author":"Sutton","year":"2019","journal-title":"npj Comput. Mater."},{"key":"mlstabdaf7bib58","first-page":"pp 8466","author":"Liu","year":"2019"},{"key":"mlstabdaf7bib59","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1038\/srep02810","volume":"3","author":"Pilania","year":"2013","journal-title":"Sci. Rep."},{"key":"mlstabdaf7bib60","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1038\/npjcompumats.2016.28","volume":"2","author":"Ward","year":"2016","journal-title":"npj Comput. Mater."},{"key":"mlstabdaf7bib61","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevMaterials.2.083802","volume":"2","author":"Ouyang","year":"2018","journal-title":"Phys. Rev. Mater."},{"key":"mlstabdaf7bib62","doi-asserted-by":"publisher","DOI":"10.1063\/1.5019779","volume":"148","author":"Sch\u00fctt","year":"2018","journal-title":"J. Chem. Phys."},{"key":"mlstabdaf7bib63","author":"Cohen","year":"2018"},{"key":"mlstabdaf7bib64","author":"Paleico","year":"2020"}],"container-title":["Machine Learning: Science and Technology"],"original-title":[],"link":[{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/abdaf7","content-type":"text\/html","content-version":"am","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/abdaf7\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/abdaf7","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/abdaf7\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/abdaf7\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"syndication"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/abdaf7\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/abdaf7\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"similarity-checking"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/abdaf7\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2022,1,11]],"date-time":"2022-01-11T09:52:29Z","timestamp":1641894749000},"score":1,"resource":{"primary":{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/abdaf7"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2021,4,14]]},"references-count":64,"journal-issue":{"issue":"2","published-online":{"date-parts":[[2021,4,14]]},"published-print":{"date-parts":[[2021,6,1]]}},"URL":"https:\/\/doi.org\/10.1088\/2632-2153\/abdaf7","relation":{},"ISSN":["2632-2153"],"issn-type":[{"value":"2632-2153","type":"electronic"}],"subject":[],"published":{"date-parts":[[2021,4,14]]},"assertion":[{"value":"The role of feature space in atomistic learning","name":"article_title","label":"Article Title"},{"value":"Machine Learning: Science and Technology","name":"journal_title","label":"Journal Title"},{"value":"paper","name":"article_type","label":"Article Type"},{"value":"\u00a9 2021 The Author(s). Published by IOP Publishing Ltd","name":"copyright_information","label":"Copyright Information"},{"value":"2020-08-26","name":"date_received","label":"Date Received","group":{"name":"publication_dates","label":"Publication dates"}},{"value":"2021-01-12","name":"date_accepted","label":"Date Accepted","group":{"name":"publication_dates","label":"Publication dates"}},{"value":"2021-04-14","name":"date_epub","label":"Online publication date","group":{"name":"publication_dates","label":"Publication dates"}}]}}