The data-intensive scientific revolution occurring where two-dimensional materials meet machine learning

Abstract

Machine learning (ML) has experienced rapid development in recent years and been widely applied to assist studies in various research areas. Two-dimensional (2D) materials, due to their unique chemical and physical properties, have been receiving increasing attention since the isolation of graphene. The combination of ML and 2D materials science has significantly accelerated the development of new functional 2D materials, and a timely review may inspire further ML-assisted 2D materials development. In this review, we provide a horizontal and vertical summary of the recent advances at the intersection of the fields of ML and 2D materials, discussing ML-assisted 2D materials preparation (design, discovery, and synthesis of 2D materials), atomistic structure analysis (structure identification and formation mechanism), and properties prediction (electronic properties, thermodynamic properties, mechanical properties, and other properties) and revealing their connections. Finally, we highlight current research challenges and provide insight into future research opportunities. The combination of machine learning and 2D materials science has significantly accelerated the development of functional 2D materials. In this review, Yin et al. provide a horizontal and vertical summary of the recent advances at the intersection of the two fields, covering ML-assisted materials discovery, atomistic structure analysis, and property prediction.