Advances in Machine Learned Potentials for Molecular Dynamics Simulation

Explore advances in machine learning for molecular dynamics simulations in this 23-minute talk by Kipton Barros from Los Alamos National Laboratory. Delve into topics such as Born Oppenheimer MD, QM9 dataset of organic molecules, levels of quantum chemistry, and symmetries of ML potentials. Learn about training techniques using energies and forces, and compare active learning to random sampling. Examine applications in hydrocarbon reaction energy benchmarks and IR spectra analysis. Gain insights into cutting-edge approaches for enhancing molecular dynamics research through machine learning techniques.

Advances in machine learned potentials for molecular dynamics simulation
Molecular dynamics articles
Born Oppenheimer MD
QM9 - 130k organic molecules Relaxed geometries
Levels of quantum chemistry
Symmetries of ML potential Translation
Training to energies and forces
Active learning vs random sampling
Hydrocarbon reaction energy benchmark
IR Spectra