Explainable machine learning in materials science

TLDR

Machine learning models are increasingly used in materials science for their high accuracy, yet the most accurate models are often difficult to interpret, prompting the emergence of explainable AI (XAI) to address this challenge. This article aims to provide an entry point to XAI for materials scientists. The authors define key XAI concepts for materials science and review example studies illustrating how XAI can aid materials research. The paper discusses the challenges and opportunities associated with applying XAI in materials science.

Abstract

Abstract Machine learning models are increasingly used in materials studies because of their exceptional accuracy. However, the most accurate machine learning models are usually difficult to explain. Remedies to this problem lie in explainable artificial intelligence (XAI), an emerging research field that addresses the explainability of complicated machine learning models like deep neural networks (DNNs). This article attempts to provide an entry point to XAI for materials scientists. Concepts are defined to clarify what explain means in the context of materials science. Example works are reviewed to show how XAI helps materials science research. Challenges and opportunities are also discussed.