Comparing machine learning techniques for predicting glassy dynamics

TL;DR

This study compares linear regression, neural networks, and GNNs in predicting glassy dynamics using advanced structural descriptors, finding that all perform similarly, but linear regression is much faster to train.

Contribution

It systematically evaluates multiple machine learning techniques with the same descriptors, highlighting the efficiency of linear regression for predicting glassy dynamics.

Findings

All three methods predict dynamics with similar accuracy.

Linear regression is significantly faster to train.

Advanced descriptors improve prediction performance.

Abstract

In the quest to understand how structure and dynamics are connected in glasses, a number of machine learning based methods have been developed that predict dynamics in supercooled liquids. These methods include both increasingly complex machine learning techniques, and increasingly sophisticated descriptors used to describe the environment around particles. In many cases, both the chosen machine learning technique and choice of structural descriptors are varied simultaneously, making it hard to quantitatively compare the performance of different machine learning approaches. Here, we use three different machine learning algorithms -- linear regression, neural networks, and GNNs -- to predict the dynamic propensity of a glassy binary hard-sphere mixture using as structural input a recursive set of order parameters recently introduced by Boattini et al. [Phys. Rev. Lett. 127, 088007 (2021)]. As we show, when these advanced descriptors are used, all three methods predict the dynamics with nearly equal accuracy. However, the linear regression is orders of magnitude faster to train making it by far the method of choice.