Improving accuracy of interatomic potentials: more physics or more data? A case study of silica
Preprint
Abstract
In this paper we test two strategies to improving the accuracy of machine-learning potentials, namely adding more fitting parameters thus making use of large volumes of available quantum-mechanical data, and adding a charge-equilibration model to account for ionic nature of the SiO2 bonding. To that end, we compare Moment Tensor Potentials (MTPs) and MTPs combined with the charge-equilibration (QEq) model (MTP+QEq) fitted to a density functional theory dataset of alpha-quartz SiO2-based structures. In order to make a meaningful comparison, in addition to the accuracy, we assess the uncertainty of predictions of each potential. It is shown that adding the QEq model to MTP does not make any improvement over the MTP potential alone, while adding more parameters does improve the accuracy and uncertainty of its predictions.Keywords
All Related Versions
This publication has 52 references indexed in Scilit:
- Accelerating high-throughput searches for new alloys with active learning of interatomic potentialsComputational Materials Science, 2019
- Neural network and ReaxFF comparison for Au propertiesInternational Journal of Quantum Chemistry, 2016
- Grand canonical molecular dynamics simulations of Cu–Au nanoalloys in thermal equilibrium using reactive ANN potentialsComputational Materials Science, 2015
- First principles phonon calculations in materials scienceScripta Materialia, 2015
- High-Dimensional Neural Network Potentials for Organic Reactions and an Improved Training AlgorithmJournal of Chemical Theory and Computation, 2015
- Spectral neighbor analysis method for automated generation of quantum-accurate interatomic potentialsJournal of Computational Physics, 2015
- Neural network‐based approaches for building high dimensional and quantum dynamics‐friendly potential energy surfacesInternational Journal of Quantum Chemistry, 2014
- Efficiency of ab-initio total energy calculations for metals and semiconductors using a plane-wave basis setComputational Materials Science, 1996
- Fast Parallel Algorithms for Short-Range Molecular DynamicsJournal of Computational Physics, 1995
- Die Berechnung optischer und elektrostatischer GitterpotentialeAnnalen der Physik, 1921