Comparing machine learning potentials for water

Author(s): Pablo Montero de Hijes, Christoph Dellago, Ryosuke Jinnouchi, Bernhard Schmiedmayer, Georg Kresse
Abstract: In this paper, we investigate the performance of different machine learning potentials (MLPs) in predicting key thermodynamic properties of water using RPBE + D3. Specifically, we scrutinize kernel-based regression and high-dimensional neural networks trained on a highly accurate dataset consisting of about 1500 structures, as well as a smaller dataset, about half the size, obtained using only on-the-fly learning. This study reveals that despite minor differences between the MLPs, their agreement on observables such as the diffusion constant and pair-correlation functions is excellent, especially for the large training dataset. Variations in the predicted density isobars, albeit somewhat larger, are also acceptable, particularly given the errors inherent to approximate density functional theory. Overall, this study emphasizes the relevance of the database over the fitting method. Finally, this study underscores the limitations of root mean square errors and the need for comprehensive testing, advocating the use of multiple MLPs for enhanced certainty, particularly when simulating complex thermodynamic properties that may not be fully captured by simpler tests.
Organisation(s): Department of Lithospheric Research, Computational and Soft Matter Physics, Computational Materials Physics
External organisation(s): Toyota Central R&D Labs., Inc., VASP Software GmbH
Journal: Journal of Chemical Physics
Volume: 160
No. of pages: 12
ISSN: 0021-9606
DOI: https://doi.org/10.1063/5.0197105
Publication date: 03-2024
Peer reviewed: Yes
Austrian Fields of Science 2012: 103043 Computational physics, 103029 Statistical physics, 102019 Machine learning
ASJC Scopus subject areas: General Physics and Astronomy, Physical and Theoretical Chemistry
Portal url: https://ucrisportal.univie.ac.at/en/publications/dcf79e5f-8275-4f4f-bf91-e2e064cdfbd1