查看文章

Machine learning applications in chemistry have been booming during the last few years, and data-driven prediction in organic synthesis witnessing a revolution. Targeting the reactions’ performances, the implementation of reaction knowledge has provided a bridge for the promotion of more sophisticated and accurate machine learning. Herein, we highlight recent, representative advances in embedding techniques and model designs with incorporated reaction-related chemical knowledge for synthetic performance predictions. For more information, see the Review by Ackermann and Hong et al.(DOI: 10.1002/chem. 202202834).