I am now full professor at the Université de Technologie de Compiègne.

My research focuses on machine learning, and more particularly on learning cautious models from scarce or poor data (Safe AI); on explainability, and in particular explaining indeterminate predictions; and on decision-making from partial information. I also work on classifier combination and data fusion issues.

I am interested in the theoretical frameworks of probability theory, belief functions (also known as Dempster-Shafer theory), and imprecise probability, for representing and managing imprecision and uncertainty.