Learning invariant representations of molecules for atomization energy prediction

Grègoire Montavon, Katja Hansen, Siamac Fazli, Matthias Rupp, Franziska Biegler, Andreas Ziehe, Alexandre Tkatchenko, O. Anatole Von Lilienfeld, Klaus Muller

Research output: Chapter in Book/Report/Conference proceedingConference contribution

52 Citations (Scopus)

Abstract

The accurate prediction of molecular energetics in chemical compound space is a crucial ingredient for rational compound design. The inherently graph-like, non-vectorial nature of molecular data gives rise to a unique and difficult machine learning problem. In this paper, we adopt a learning-from-scratch approach where quantum-mechanical molecular energies are predicted directly from the raw molecular geometry. The study suggests a benefit from setting flexible priors and enforcing invariance stochastically rather than structurally. Our results improve the state-of-the-art by a factor of almost three, bringing statistical methods one step closer to chemical accuracy.

Original languageEnglish
Title of host publicationAdvances in Neural Information Processing Systems
Pages440-448
Number of pages9
Volume1
Publication statusPublished - 2012 Dec 1
Event26th Annual Conference on Neural Information Processing Systems 2012, NIPS 2012 - Lake Tahoe, NV, United States
Duration: 2012 Dec 32012 Dec 6

Other

Other26th Annual Conference on Neural Information Processing Systems 2012, NIPS 2012
CountryUnited States
CityLake Tahoe, NV
Period12/12/312/12/6

Fingerprint

Chemical compounds
Atomization
Invariance
Learning systems
Statistical methods
Molecules
Geometry

ASJC Scopus subject areas

  • Computer Networks and Communications
  • Information Systems
  • Signal Processing

Cite this

Montavon, G., Hansen, K., Fazli, S., Rupp, M., Biegler, F., Ziehe, A., ... Muller, K. (2012). Learning invariant representations of molecules for atomization energy prediction. In Advances in Neural Information Processing Systems (Vol. 1, pp. 440-448)

Learning invariant representations of molecules for atomization energy prediction. / Montavon, Grègoire; Hansen, Katja; Fazli, Siamac; Rupp, Matthias; Biegler, Franziska; Ziehe, Andreas; Tkatchenko, Alexandre; Von Lilienfeld, O. Anatole; Muller, Klaus.

Advances in Neural Information Processing Systems. Vol. 1 2012. p. 440-448.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Montavon, G, Hansen, K, Fazli, S, Rupp, M, Biegler, F, Ziehe, A, Tkatchenko, A, Von Lilienfeld, OA & Muller, K 2012, Learning invariant representations of molecules for atomization energy prediction. in Advances in Neural Information Processing Systems. vol. 1, pp. 440-448, 26th Annual Conference on Neural Information Processing Systems 2012, NIPS 2012, Lake Tahoe, NV, United States, 12/12/3.
Montavon G, Hansen K, Fazli S, Rupp M, Biegler F, Ziehe A et al. Learning invariant representations of molecules for atomization energy prediction. In Advances in Neural Information Processing Systems. Vol. 1. 2012. p. 440-448
Montavon, Grègoire ; Hansen, Katja ; Fazli, Siamac ; Rupp, Matthias ; Biegler, Franziska ; Ziehe, Andreas ; Tkatchenko, Alexandre ; Von Lilienfeld, O. Anatole ; Muller, Klaus. / Learning invariant representations of molecules for atomization energy prediction. Advances in Neural Information Processing Systems. Vol. 1 2012. pp. 440-448
@inproceedings{02648bbe33a647e5b2a6034fe97ccd1f,
title = "Learning invariant representations of molecules for atomization energy prediction",
abstract = "The accurate prediction of molecular energetics in chemical compound space is a crucial ingredient for rational compound design. The inherently graph-like, non-vectorial nature of molecular data gives rise to a unique and difficult machine learning problem. In this paper, we adopt a learning-from-scratch approach where quantum-mechanical molecular energies are predicted directly from the raw molecular geometry. The study suggests a benefit from setting flexible priors and enforcing invariance stochastically rather than structurally. Our results improve the state-of-the-art by a factor of almost three, bringing statistical methods one step closer to chemical accuracy.",
author = "Gr{\`e}goire Montavon and Katja Hansen and Siamac Fazli and Matthias Rupp and Franziska Biegler and Andreas Ziehe and Alexandre Tkatchenko and {Von Lilienfeld}, {O. Anatole} and Klaus Muller",
year = "2012",
month = "12",
day = "1",
language = "English",
isbn = "9781627480031",
volume = "1",
pages = "440--448",
booktitle = "Advances in Neural Information Processing Systems",

}

TY - GEN

T1 - Learning invariant representations of molecules for atomization energy prediction

AU - Montavon, Grègoire

AU - Hansen, Katja

AU - Fazli, Siamac

AU - Rupp, Matthias

AU - Biegler, Franziska

AU - Ziehe, Andreas

AU - Tkatchenko, Alexandre

AU - Von Lilienfeld, O. Anatole

AU - Muller, Klaus

PY - 2012/12/1

Y1 - 2012/12/1

N2 - The accurate prediction of molecular energetics in chemical compound space is a crucial ingredient for rational compound design. The inherently graph-like, non-vectorial nature of molecular data gives rise to a unique and difficult machine learning problem. In this paper, we adopt a learning-from-scratch approach where quantum-mechanical molecular energies are predicted directly from the raw molecular geometry. The study suggests a benefit from setting flexible priors and enforcing invariance stochastically rather than structurally. Our results improve the state-of-the-art by a factor of almost three, bringing statistical methods one step closer to chemical accuracy.

AB - The accurate prediction of molecular energetics in chemical compound space is a crucial ingredient for rational compound design. The inherently graph-like, non-vectorial nature of molecular data gives rise to a unique and difficult machine learning problem. In this paper, we adopt a learning-from-scratch approach where quantum-mechanical molecular energies are predicted directly from the raw molecular geometry. The study suggests a benefit from setting flexible priors and enforcing invariance stochastically rather than structurally. Our results improve the state-of-the-art by a factor of almost three, bringing statistical methods one step closer to chemical accuracy.

UR - http://www.scopus.com/inward/record.url?scp=84877770101&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84877770101&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:84877770101

SN - 9781627480031

VL - 1

SP - 440

EP - 448

BT - Advances in Neural Information Processing Systems

ER -