Quantum-Chemical Insights from Interpretable Atomistic Neural Networks

dc.contributor.authorSchütt, Kristof T.
dc.contributor.authorGastegger, Michael
dc.contributor.authorTkatchenko, Alexandre
dc.contributor.authorMüller, Klaus-Robert
dc.date.accessioned2020-06-19T08:06:56Z
dc.date.available2020-06-19T08:06:56Z
dc.date.issued2019-09-10
dc.description.abstractWith the rise of deep neural networks for quantum chemistry applications, there is a pressing need for architectures that, beyond delivering accurate predictions of chemical properties, are readily interpretable by researchers. Here, we describe interpretation techniques for atomistic neural networks on the example of Behler–Parrinello networks as well as the end-to-end model SchNet. Both models obtain predictions of chemical properties by aggregating atom-wise contributions. These latent variables can serve as local explanations of a prediction and are obtained during training without additional cost. Due to their correspondence to well-known chemical concepts such as atomic energies and partial charges, these atom-wise explanations enable insights not only about the model but more importantly about the underlying quantum-chemical regularities. We generalize from atomistic explanations to 3d space, thus obtaining spatially resolved visualizations which further improve interpretability. Finally, we analyze learned embeddings of chemical elements that exhibit a partial ordering that resembles the order of the periodic table. As the examined neural networks show excellent agreement with chemical knowledge, the presented techniques open up new venues for data-driven research in chemistry, physics and materials science.en
dc.description.sponsorshipEC/H2020/792572/EU/Machine Learning for Catalytic Carbon Dioxide Activation/MachineCaten
dc.description.sponsorshipBMBF, 01IS14013A, Verbundprojekt: BBDC - Berliner Kompetenzzentrum für Big Dataen
dc.description.sponsorshipBMBF, 01IS18037A, Verbundprojekt BIFOLD-BZML: Berlin Institute for the Foundations of Learning and Dataen
dc.description.sponsorshipEC/H2020/725291/EU/Beyond Static Molecules: Modeling Quantum Fluctuations in Complex Molecular Environments/BeStMoen
dc.identifier.eissn1611-3349
dc.identifier.isbn978-3-030-28954-6
dc.identifier.isbn978-3-030-28953-9
dc.identifier.issn0302-9743
dc.identifier.urihttps://depositonce.tu-berlin.de/handle/11303/11437
dc.identifier.urihttp://dx.doi.org/10.14279/depositonce-10318
dc.language.isoenen
dc.rights.urihttp://rightsstatements.org/vocab/InC/1.0/en
dc.subject.ddc541 Physikalische Chemiede
dc.subject.ddc006 Spezielle Computerverfahrende
dc.subject.othermachine learningen
dc.subject.otherchemistryen
dc.subject.otherexplanationen
dc.titleQuantum-Chemical Insights from Interpretable Atomistic Neural Networksen
dc.typeBook Parten
dc.type.versionacceptedVersionen
dcterms.bibliographicCitation.booktitleExplainable AI: Interpreting, Explaining and Visualizing Deep Learningen
dcterms.bibliographicCitation.doi10.1007/978-3-030-28954-6_17en
dcterms.bibliographicCitation.editorSamek, Wojciech
dcterms.bibliographicCitation.editorMontavon, Grégoire
dcterms.bibliographicCitation.editorVedaldi, Andrea
dcterms.bibliographicCitation.editorHansen, Lars Kai
dcterms.bibliographicCitation.editorMüller, Klaus-Robert
dcterms.bibliographicCitation.originalpublishernameSpringeren
dcterms.bibliographicCitation.originalpublisherplaceChamen
dcterms.bibliographicCitation.pageend330en
dcterms.bibliographicCitation.pagestart311en
tub.accessrights.dnbfreeen
tub.affiliationFak. 4 Elektrotechnik und Informatik>Inst. Softwaretechnik und Theoretische Informatik>FG Maschinelles Lernende
tub.affiliation.facultyFak. 4 Elektrotechnik und Informatikde
tub.affiliation.groupFG Maschinelles Lernende
tub.affiliation.instituteInst. Softwaretechnik und Theoretische Informatikde
tub.publisher.universityorinstitutionTechnische Universität Berlinen
tub.series.issuenumber11700en
tub.series.nameLecture Notes in Computer Scienceen
Files
Original bundle
Now showing 1 - 1 of 1
Loading…
Thumbnail Image
Name:
schuett_etal_2019.pdf
Size:
4.4 MB
Format:
Adobe Portable Document Format
Description:
Accepted manuscript
Collections