Predicting materials properties without crystal structure: Deep representation learning from stoichiometry

TL;DR

This paper introduces a deep learning method that predicts materials properties using only stoichiometry, automatically learning effective representations without needing crystal structure data, thus broadening applicability and improving accuracy.

Contribution

It presents a novel graph-based deep learning approach that learns descriptors directly from stoichiometry, outperforming existing structure-agnostic methods.

Findings

Lower prediction errors than state-of-the-art methods

Requires less data for training

Effective for materials without known crystal structures

Abstract

Machine learning has the potential to accelerate materials discovery by accurately predicting materials properties at a low computational cost. However, the model inputs remain a key stumbling block. Current methods typically use descriptors constructed from knowledge of either the full crystal structure -- therefore only applicable to materials with already characterised structures -- or structure-agnostic fixed-length representations hand-engineered from the stoichiometry. We develop a machine learning approach that takes only the stoichiometry as input and automatically learns appropriate and systematically improvable descriptors from data. Our key insight is to treat the stoichiometric formula as a dense weighted graph between elements. Compared to the state of the art for structure-agnostic methods, our approach achieves lower errors with less data.