Computational chemistry and AI
The course “Computational chemistry and AI” teaches essential skills in applying AI for molecular simulations at the quantum chemical levels. It consists of plenty of hands-on exercises going along with the fundamental theoretical background. The course will introduce state-of-the-art machine-learning (ML) approaches for accelerating and improving accuracy of quantum chemical simulations. It will also use the established program for such simulations, MLatom, providing an excellent tool set for both complete beginners and advanced students. Many of the exercises can be conviniently performed online at Aitomistic Lab@XMU. At the end of the course, the students will be able to use the quantum chemical methods and universal ML models for their simulations including geometry optimizations, thermochemistry calculations, ground- and excited-state molecular dynamics, various types of spectroscopy (IR, Raman, UV/vis, two-photon absorption), explore reaction mechanisms as well as create their own ML interatomic potentials and custom models, and build their own data sets via active learning approaches.
An essential, brief non-mathematical introductions to the background knowledge of quantum chemistry and molecular simulations are provided too. The knowledge of Python is a plus but not required. No preliminary knowledge of ML is necessary.