(657i) Physically-Motivated Requirements of Machine Learning Potentials | AIChE

(657i) Physically-Motivated Requirements of Machine Learning Potentials


Mason, J., UC Davis
Machine learning potentials (MLPs) for molecular dynamics simulations have been found to be capable of approaching ab initio accuracy with the computational efficiency of empirical potentials. The accuracy and performance of an MLP is highly dependent upon the choice of descriptors used to characterize the local atomic environment though, as well as on the training data that the algorithm uses to predict the energies and forces. The descriptors should respect the symmetries of the atomic environment, be differentiable with respect to the atomic coordinates, and contain sufficient information. If a Gaussian process is used for machine learning, the covariance function that quantifies the similarity or closeness between two points in descriptor space also has requirements to consider. Specifically, the choice of covariance function should impose minimal constraints on the potential to reduce the risk of systematic error. This work reports on recent progress using Gaussian process regression and a novel set of descriptors to construct accurate and efficient interatomic potentials.