Research Interest

During my Ph.D. at the DAVID laboratory, I first focused on scaling MCMC inference algorithms for Dirichlet Process Mixture Models applied to clustering, as well as on Bayesian nonparametric latent block models for co-clustering. In a second phase, I studied the stability of the EM algorithm applied to spherical Gaussian mixture models.

My current research now develops along two complementary directions. The first aims to extend and deepen the topics explored during my Ph.D., while the second addresses broader questions in statistical learning and deep learning, particularly the regularization of neural networks to improve their generalization and robustness.