Machine learning

Machine learning has entered the field of quantum matter with applications covering quantum materials and the many-body problem. Interpretable and computationally-efficient machine learning models are able to capture the structure-property relationship in materials science.

Among others, we use the organic materials database developed within our group as a training set for our machine learning studies. The database hosts electronic and magnetic structures of about 25,000 3-dimensional organic crystals and provides a highly complex dataset to work on. Applying machine learning we intend to provide predictions towards novel functional materials based on the properties calculated within our training sets.

Key Papers:
  • Band gap prediction for large organic crystal structures with machine learning
    Bart Olsthoorn, R. Matthias Geilhufe, Stanislav S. Borysov, Alexander V. Balatsky