Research focuses on the theoretical foundations of probability, statistical estimation, Bayesian and frequentist inference methods, and Monte Carlo algorithms (sequential and MCMC). The research also includes developing and analyzing modern variational inference methods, which are used to efficiently approximate complex distributions.
Particular attention is given to modeling and analyzing complex unpredictable phenomena, including in the presence of constraints, such as those encountered in reflected Brownian motion, algorithm convergence, and applications in signal processing, imaging, and data science.
Members:
- Pierre Le Bris
- Randal Douc
- Xavier Erny
- Sandro Franceschi
- Hélène Halconruy (group leader)
- Wojciech Pieczynski (Emeritus)
