I currently work on application of machine learning, and in particular unsupervised learning to material science and chemistry. In our everyday life, we rely on a relatively small number of materials, with very specific characteristics: ceramics, metallic alloys, polymers, etc. However, the number of possible materials created by combining different chemical elements in various proportions and arrangements is gigantic. Because of this, exploring the universe of possible materials to find the best one for a given application is a daunting task, which can be helped by using dimensionality reduction techniques and visualization algorithms.
I am particularly interested in devising, implementing and studying algorithms used to simulate complex chemical behavior of simple systems and emerging properties.
I did my Ph.D. with François-Xavier Coudert at Chimie ParisTech — PSL University, on the coupling of intrusion and adsorption to the deformations of flexible nanoporous materials. I used simulation techniques ranging from ab initio dynamics to classical free energy methods and thermodynamic calculations to improve our understanding of the atomistic processes in theses systems.