Improving accuracy of interatomic potentials: more physics or more data? A case study of silica
arXiv:1808.03783
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.
References in corpus (8)
- Neural Message Passing for Quantum Chemistry
- A Spectral Analysis Method for Automated Generation of Quantum-Accurate Interatomic Potentials
- Comparing molecules and solids across structural and alchemical space
- Machine learning prediction errors better than DFT accuracy
- Machine-learning based interatomic potential for amorphous carbon
- Symmetry-Adapted Machine-Learning for Tensorial Properties of Atomistic Systems
- Data-driven learning of total and local energies in elemental boron
- A learning scheme to predict atomic forces and accelerate materials simulations