Combining Machine Learning Potential and Structure Prediction for Accelerated Materials Design and Discovery

Abstract

The theoretical structure prediction method via quantum mechanical atomistic simulations such as density functional theory (DFT), based solely on chemical composition, has already become a routine tool to determine the structures of physical and chemical systems, e.g., solids and clusters. However, the application of DFT to more realistic simulations, to a large extent, is impeded because of the unfavorable scaling of the computational cost with respect to the system size. During recent years, the machine learning potential (MLP) method has been rapidly rising as an accurate and efficient tool for atomistic simulations. In this Perspective, we provide an introduction to the basic principles and advantages of the combination of structure prediction and MLP, as well as the challenges and opportunities associated with this promising approach.

Gaussian Process Regression for Materials and Molecules

“…Here, we focus the presentation on GAP, in keeping with the scope of the present review article, but we note that other ML fitting schemes have also been successfully combined with different structure-searching techniques. 126 , 134 , 141 …”

“… 134 We focus on GAP below, but we note that more generally, the ways in which crystal structure prediction can be accelerated using machine-learned force fields (including various fitting schemes and their applications) have been reviewed in a recent perspective article. 141 …”

Choosing the right molecular machine learning potential

“…Recently, machine learning (ML) has emerged as a promising approach that is rocking the foundations of how we simulate molecular PES. [2][3][4][5][6][7][8][9] Built on statistical principles, MLbased PESs, or more simply ML potentials (MLPs), aim to identify an unbiased predicting function that optimally correlates a set of molecular structures with the given target energies and, often, forces used as training data. (The force acting on the nuclei is the negative of the PES gradient.)…”

Choosing the right molecular machine learning potential

“…(The force acting on the nuclei is the negative of the PES gradient.) Owing to its generalization capabilities and fast prediction on unseen data, MLPs can be explored to accelerate minimum-energy [10][11][12][13][14][15] and transition-state structure search, 13,16,17 vibrational analysis, 18-21 absorption 22,23 and emission spectra simulation, 24 reaction 13, 25,26 and structural transition exploration, 27 and ground- [3][4][5][6][7][8][9] and excitedstate dynamics propagation. 28,29 A blessing and a curse of ML is that it is possible to design, for all practical purposes, an infinite number of MLP models that can describe a molecular PES.…”

“…Thus, various reviews, perspectives, tutorials, and books have been published at an ever-increasing pace to survey state-of-the-art MLPs (just a small selection of reviews are in Refs. [3][4][5][6][7][8][9]. Complementary to these studies, an effort was made to benchmark 40 the performance (accuracy and efficiency) of MLPs with respect to energy predictions by focusing specifically on the algorithm component of the models.…”