Modeling of materials and its applications using density functional theory calculation and machine learning

Abstract

This paper demonstrates machine learning potential, providing an accurate description of the relationship between the energy and the crystal structure and its potential applications such as large-scale molecular dynamics simulations and global structure optimizations. This paper also describes machine-learning-based methods to perform crystal structure optimization from a vast number of candidate structures. Finally, the paper introduces a novel compact data structure of the zero-suppressed binary decision diagram (ZDD) to enumerate structures much more efficiently.