New paper in Chemical Science
A transferable active-learning strategy for reactive molecular force fields
Published in Chemical Science, TMCS CDT students Tom Young and Tristan Johnston-Wood have developed an efficient strategy for training machine learning (ML) models for the approximation of potential energy surfaces. The method and code are open source and completely automated, enabling the efficient generation of accurate ML potentials to study chemical reactions. Compared to conventional quantum mechanics approaches, this strategy enables faster and more routine interrogation of dynamic effects in organic and organometallic reactions.
