Organic Reaction Mechanism Classification Using Machine Learning
Igor Larrosa and Jordi Burés describe the application of machine learning in understanding reaction mechanisms.
Mechanistic studies are a type of mathematical inverse problem in which chemists analyze a set of observations (kinetic data) to uncover the underlying causes (reaction mechanisms). According to Professor Igor Larrosa from the University of Manchester (UK) a modern approach to solving these problems involves training deep neural networks with known data. He said that “this can be compared to exposing the human brain to a large quantity of kinetic data generated by each mechanism, allowing it to learn the intrinsic patterns and develop the ability to deduce the reaction mechanism from previously unseen kinetic data.” Professor Larrosa and Dr. Jordi Burés, also from the University of Manchester, strongly believe in the potential of using mechanistic insights to enhance the discovery and improvement of catalytic reactions.
Read the full article Organic Reaction Mechanism Classification Using Machine Learning