seminar on Machine learned interatomic potentials

The Multi-Scale Mechanics (MSM) group (PI: Dr. Francesco Maresca) presents:

Machine learned interatomic potentials

Prof. Gábor Csányi, University of Cambridge (UK)

Abstract: I will make the somewhat bold claim that over the past 10 years, a new computational task has been defined and solved for extended material systems: this is the analytic fitting of the Born-Oppenheimer potential energy surface as a function of nuclear coordinates under the assumption of medium-range interactions, out to 5-10 Å. The resulting potentials are reactive, many-body, reach accuracies of a few meV/atom, with costs that are on the order of 1-10 ms/atom. Important challenges remain: treatment of long range interactions in a nontrivial way, e.g. environment dependent multipoles, charge transfer, magnetism. Time is ripe for a “shakedown” of the details among various approaches (neural networks, kernels, polynomials), and more standard protocols of putting together the training data. Tradeoffs between system- (or even project-) specific fits vs. more general potentials will be ongoing. I am particularly concerned with the amount of physics and chemistry that we impute into these approximations, and that can be used to help "extrapolate" correctly into regions of configuration space far from those in the data set.

Bio: Dr. Gábor Csányi is Professor of Molecular Modelling at the Department of Engineering of the University of Cambridge. He is expert in atomistic simulation, particularly in multiscale modelling that couples quantum mechanics to larger length scales. He is currently engaged in applying machine learning and other data intensive techniques to materials modelling problems, such as deriving force fields (interatomic potentials) from ab initio electronic structure data.