Research Projects

The main research theme of the group is an application of machine learning to molecular modeling. The projects that we currently work on are:

• Machine-learning interatomic potentials. These are models of interaction of atoms that are alternative to quantum-mechanical models (too slow) and empirical potentials (too inaccurate). We collaborate with materials scientists and chemists on:
  • materials discovery
  • first-principles calculation of alloy phase diagrams
  • prediction of chemical reaction rates