Congrats to Hao and Yitian on the acceptance of short-for-long machine-learning for nonadiabatic dynamics by J. Chem. Theory Comput.! This paper introduces a practical way to estimate memory time in non-Markovian dynamics of reduced density matrix and also compares several ML methods and physics-based methods.
- Hao Zeng, Yitian Kou, Xiang Sun*, How Sophisticated Are Neural Networks Needed to Predict Long-Term Nonadiabatic Dynamics? J. Chem. Theory Comput. (2024) https://pubs.acs.org/doi/10.1021/acs.jctc.4c01223