Learning the Exchange-Correlation Functional from Nature with Fully Differentiable Density Functional Theory

M. F. Kasim and S. M. Vinko
Phys. Rev. Lett. 127, 126403 – Published 15 September 2021
PDFHTMLExport Citation

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 modeling 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 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.

  • Figure
  • Figure
  • Figure
  • Received 8 February 2021
  • Accepted 17 August 2021

DOI:https://doi.org/10.1103/PhysRevLett.127.126403

© 2021 American Physical Society

Physics Subject Headings (PhySH)

Condensed Matter, Materials & Applied Physics

Authors & Affiliations

M. F. Kasim1,* and S. M. Vinko1,2,†

  • 1Department of Physics, Clarendon Laboratory, University of Oxford, Parks Road, Oxford OX1 3PU, United Kingdom
  • 2Central Laser Facility, STFC Rutherford Appleton Laboratory, Didcot OX11 0QX, United Kingdom

  • *muhammad.kasim@physics.ox.ac.uk Present address: Machine Discovery, Oxford OX4 4GP, United Kingdom.
  • sam.vinko@physics.ox.ac.uk

Article Text (Subscription Required)

Click to Expand

Supplemental Material (Subscription Required)

Click to Expand

References (Subscription Required)

Click to Expand
Issue

Vol. 127, Iss. 12 — 17 September 2021

Reuse & Permissions
Access Options
Author publication services for translation and copyediting assistance advertisement

Authorization Required


×
×

Images

×

Sign up to receive regular email alerts from Physical Review Letters

Log In

Cancel
×

Search


Article Lookup

Paste a citation or DOI

Enter a citation
×