Learning inducing points and uncertainty on molecular data by scalable variational Gaussian processes

TL;DR

This paper demonstrates that scalable variational Gaussian processes with learned inducing points improve molecular energy and force predictions, providing reliable uncertainty estimates and generalizing to unseen molecular configurations.

Contribution

It introduces a variational learning approach for inducing points in molecular data, enhancing prediction accuracy and uncertainty quantification in large-scale molecular datasets.

Findings

Variational GPs improve energy and force predictions on molecular datasets.

Predictive log-likelihood yields better uncertainty estimates.

Models generalize to unseen molecular configurations.

Abstract

Uncertainty control and scalability to large datasets are the two main issues for the deployment of Gaussian process (GP) models within the autonomous machine learning-based prediction pipelines in material science and chemistry. One way to address both of these issues is by introducing the latent inducing point variables and choosing the right approximation for the marginal log-likelihood objective. Here, we empirically show that variational learning of the inducing points in a molecular descriptor space improves the prediction of energies and atomic forces on two molecular dynamics datasets. First, we show that variational GPs can learn to represent the configurations of the molecules of different types that were not present within the initialization set of configurations. We provide a comparison of alternative log-likelihood training objectives and variational distributions. Among several evaluated approximate marginal log-likelihood objectives, we show that predictive log-likelihood provides excellent uncertainty estimates at the slight expense of predictive quality. Furthermore, we extend our study to a large molecular crystal system, showing that variational GP models perform well for predicting atomic forces by efficiently learning a sparse representation of the dataset.