Active learning of linearly parametrized interatomic potentials

2016/11/30 by Evgeny V. Podryabinkin, Alexander V. Shapeev · 2 citations
Physics and Astronomy · #physics.comp-ph #cond-mat.mtrl-sci

paper · pdf · doi:10.1016/j.commatsci.2017.08.031

published as Computational Materials Science, volume 140, pages 171-180, 2017 · 12 pages, numerical tests added

arxiv created 2017/08/22 · arxiv updated 2017/09/19

Abstract

This paper introduces an active learning approach to the fitting of machine learning interatomic potentials. Our approach is based on the D-optimality criterion for selecting atomic configurations on which the potential is fitted. It is shown that the proposed active learning approach is highly efficient in training potentials on the fly, ensuring that no extrapolation is attempted and leading to a completely reliable atomistic simulation without any significant decrease in accuracy. We apply our approach to molecular dynamics and structure relaxation, and we argue that it can be applied, in principle, to any other type of atomistic simulation. The software, test cases, and examples of usage are published at http://gitlab.skoltech.ru/shapeev/mlip/.

Cited by

• Machine-Learned Interatomic Potentials for Predicting Physicochemical Properties of Molten Metal-Salt Systems for Calcium Electrolysis
• Moment Tensor Potential and Equivariant Tensor Network Potential with explicit dispersion interactions