Abstract:

In recent years, machine learning has demonstrated great potential in chemistry. More specifically, it has found important applications in the fields of chemical property prediction, reaction performance prediction, molecular design, retrosynthetic analysis, and robotic synthesis. This paper introduces the research background, methods, prospects, and challenges of machine-learning applications in chemistry from an anthropomorphic perspective of paper A.

Key words: Machine learning, Molecular descriptors, Chemical reaction prediction, Chemical property prediction, Popularization of science