Representations of molecules and materials for interpolation of quantum-mechanical simulations via machine learning
This addresses the cost problem for researchers in physics, chemistry, and materials science, but it is incremental as it reviews and compares existing methods rather than introducing new ones.
The paper tackles the high computational cost of quantum-mechanical simulations by using machine learning for interpolation, requiring effective molecular and material representations, and it reviews and compares state-of-the-art representations in controlled experiments.
Computational study of molecules and materials from first principles is a cornerstone of physics, chemistry, and materials science, but limited by the cost of accurate and precise simulations. In settings involving many simulations, machine learning can reduce these costs, often by orders of magnitude, by interpolating between reference simulations. This requires representations that describe any molecule or material and support interpolation. We comprehensively review and discuss current representations and relations between them, using a unified mathematical framework based on many-body functions, group averaging, and tensor products. For selected state-of-the-art representations, we compare energy predictions for organic molecules, binary alloys, and Al-Ga-In sesquioxides in numerical experiments controlled for data distribution, regression method, and hyper-parameter optimization.