Colloquium J. Behler

Understanding Complex Chemical Systems with Neural Network Potentials

Prof. Jörg Behler, Göttingen University,

Thursday 16 May, 14h00, Amphi. Herpin (Bâtiment Esclangon).

A lot of progress has been made in recent years in the development of atomistic potentials employing machine learning (ML). In contrast to most conventional potentials, which are based on physical approximations to derive an analytic functional relation between the atomic configuration and the potential-energy, ML potentials rely on simple but very flexible mathematical terms without a direct physical meaning. Instead, in case of ML potentials the topology of the potential- energy surface is “learned” by adjusting a number of parameters with the aim to reproduce a set of reference electronic structure data as
accurately as possible. Due to this bias-free construction they are applicable to a wide range of systems without changes in their functional form, and a very high accuracy close to the underlying first-principles data can be obtained.

Neural network potentials (NNPs), which have first been proposed about two decades ago, are an important class of ML potentials. While the first NNPs have been restricted to small molecules with only a few degrees of freedom, they are now applicable to high-dimensional systems containing thousands of atoms, which enables addressing a variety of problems in chemistry, physics and materials science. In this talk the basic ideas of NNPs are presented with a special focus on constructing NNPs for high-dimensional condensed systems. Applications for different types of systems, from bulk materials via liquid water to processes at interfaces are presented.