Speaker: Lenz Fiedler (Helmholtz-Zentrum Dresden-Rossendorf)
Session Chair: Michael Herbst (EPFL)

Abstract: Quantum mechanical calculations of the electronic structure of matter enable accessing interesting thermodynamical properties without the need for prior experimental measurements. Therefore, electronic structure calculations are of great interest in fields such as materials discovery or drug design. At the forefront of such simulations lies density functional theory (DFT), due to its excellent balance between computational accuracy and efficiency. Yet, as pressing environmental and social issues shift the research focus to increasingly complicated systems and conditions, even the most efficient of DFT implementations are approaching their limitations in terms of computational feasibility. A possible route to enable more complex calculations lies with machine learning (ML), i.e., algorithms that are capable of capturing complicated relationships based on large amounts of data.
In this talk, Lenz Fiedler will talk about current contributions of Center for Advanced Systems Understanding, Helmholtz-Zentrum Dresden-Rossendorf (CASUS) w.r.t. building ML models that replace conventional DFT calculations. More precisely, Lenz will talk about the current state of the Materials Learning Algorithms library (MALA), which allows easy training and inference for ML-DFT models that are developed by CASUS in cooperation with Sandia National Laboraties and Oak Ridge National Laboratory. In contrast to comparable frameworks, MALA allows full access to the electronic structure of compounds, including volumetric data as well as scalar quantities of interest, such as energies. It will be shown how MALA models can operate efficiently across phase boundaries, length scales and temperature ranges.