Integrating Machine Learning and Molecular Simulation for Material Design and Discovery

Machine learning (ML) and artificial intelligence (AI) have enabled transformative impact on materials science by accelerating cutting-edge insights from computational methods and their analysis to hitherto unattainable scales. Such an assembly of linear algebra and statistical methods can facilitate the conceptual development of flexible techniques by finding mechanism/information/hidden pattern in a data set. The present review provides basic information about the classification of ML methodology and its workflow. These sections also elaborate on the advantages and limitations of various ML algorithms for solving problems in materials science and reviewing cases of success and failure. Subsequently, we show how these techniques can uncover the complexities in several quantitative structure–property relationships to design and discover novel materials for various applications. We conclude our review with an outlook on present research challenges, problems, and potential future perspectives in the field of machine learning. Overall, this review can serve as a fundamental guide to amplify the adoption of such tools and methods by materials scientists across academia and industry.