Application and Prospect of Machine Learning in the Field of Materials and Chemical Engineering

Abstract

With the rapid development of computer science, machine learning has gradually penetrated into various research fields and led to the development of related fields to different degrees. At recent stage, machine learning (ML) algorithms have been able to extensively search for variable relationships from a large amount of observed data, thus helping people to discover more laws that were once unnoticed. As an emerging approach, ML can efficiently utilize and process large-volume data generated from high-throughput experiments in the field of materials and chemicals. In recent years, with the continuous improvement of various algorithmic tools, more and more researchers have applied ML to materials and chemical research, which has greatly broadened the research ideas and methods. This paper briefly introduces the development of machine learning in recent years; summarizes the classification methods of common molecular descriptors and their types; introduces the classification, common algorithms and application scenarios of different types of ML; reviews the relevant research results in the field of materials prediction and inverse synthesis analysis of machine learning; puts forward related suggestions for the application of ML for research; and finally summarizes the prospects for the future development of machine learning applied to the field of materials and chemical engineering.