Efficient Calculations of Electronic Structures with Machine-Learning Models

Explore a comprehensive lecture on the application of machine learning models for efficient electronic structure calculations. Delve into the world of quantum mechanical simulations and density functional theory (DFT) as Lenz Fiedler from Helmholtz-Zentrum Dresden-Rossendorf discusses their importance in materials discovery and drug design. Learn about the challenges faced by conventional DFT implementations when dealing with complex systems and discover how machine learning algorithms offer a promising solution. Gain insights into the Materials Learning Algorithms library (MALA), developed by CASUS in collaboration with Sandia National Laboratories and Oak Ridge National Laboratory, which enables easy training and inference for ML-DFT models. Understand how MALA provides full access to electronic structures, including volumetric data and scalar quantities, and how its models can efficiently operate across phase boundaries, length scales, and temperature ranges.