Training-free hyperparameter optimization of neural networks for electronic structures in matter

TL;DR

This paper presents a training-free approach to hyperparameter optimization for neural networks used in electronic structure calculations, significantly reducing computational costs in materials science simulations.

Contribution

It introduces a method to bypass extensive training during hyperparameter tuning, accelerating neural network model development for quantum simulations.

Findings

Achieved roughly 100x speed-up in hyperparameter optimization

Demonstrated effectiveness on Kohn-Sham density functional theory

Reduced computational overhead in electronic structure modeling

Abstract

A myriad of phenomena in materials science and chemistry rely on quantum-level simulations of the electronic structure in matter. While moving to larger length and time scales has been a pressing issue for decades, such large-scale electronic structure calculations are still challenging despite modern software approaches and advances in high-performance computing. The silver lining in this regard is the use of machine learning to accelerate electronic structure calculations -- this line of research has recently gained growing attention. The grand challenge therein is finding a suitable machine-learning model during a process called hyperparameter optimization. This, however, causes a massive computational overhead in addition to that of data generation. We accelerate the construction of neural network models by roughly two orders of magnitude by circumventing excessive training during the hyperparameter optimization phase. We demonstrate our workflow for Kohn-Sham density functional theory, the most popular computational method in materials science and chemistry.