Talk by Kelin Xia

Artificial intelligence (AI) based molecular data analysis has begun to gain momentum due to the great advancement in experimental data, computational power and learning models. However, a major issue that remains for all AI-based learning models is the efficient molecular representations and featurization. Here we propose persistent Dirac models for the description and characterization of molecular structures and interactions. In our persistent Dirac model, molecules are represented as simplicial complexes and a series of Dirac matrices are constructed from the filtration process. Persistence attributes, which characterize the variation and persistence of eigen spectrum information from the Dirac matrices, are used as molecular descriptors or fingerprints. These features are combined with machine learning and used in material data analysis, in particular, the study of halide perovskites.