I am interested in computational statistics, generally in problems at the frontier of computer science and statistics with applications in biology.

Clustering
I have mainly worked on clustering problems. In particular I am interested in statistics on objects having a particular structure (hypersphere for transcriptomics, simplex for amino-acids distribution, contingency tables). The more usual application I have in mind is clustering with mixture models.  A current direction of research I am exploring is to adapt mixture models so that they would find non-convex clusters as the one which are found in the bull's eye example. Besides, I work with Alejandro Murua on a related problem : clustering with the Potts model.

Supervised learning
This project is complementary with the first and deals with supervised learning. I am interested in the practical aspects of supervised learning but recently started also to gain interest in the theoretical aspects of machine learning. A big goal would be to understand better how deep learning is working.

Data analysis
I have always been interested in data analysis, old school data analysis without any probabilistic model. I like particularly understanding the structure of data.

Sampling
An interest lying aside from data analysis is random sampling, the idea is to explore randomly interesting structures like contingency tables, Latin squares,