Alain Durmus

Alain Durmus is a Professor at École Polytechnique, affiliated with the CMAP (Applied Mathematics Department). His research focuses on computational statistics, machine learning, and stochastic methods for Bayesian inference and generative models. Key areas include Monte Carlo methods, stochastic optimization, and Riemannian geometry in optimization. He has contributed to theoretical advancements in Markov processes and their applications in sampling algorithms.

• Affiliations: École Polytechnique (CMAP), ENS Paris-Saclay (teaching roles), and other academic collaborations.
• Teaching: Courses on mathematical statistics, stochastic methods, probability theory, and optimization at ENS Paris-Saclay and École Polytechnique.
• Research Interests: Development of Monte Carlo algorithms, stochastic approximation schemes, and theoretical analysis of Markov chain convergence. Specializes in Bayesian statistics, generative models, and high-dimensional inference.

His work bridges theory and application, with notable contributions to Langevin dynamics, nonreversible MCMC methods, and federated learning. He has received a Best student paper award at ICASSP 2020 for collaborative research on Bayesian computation using Sliced-Wasserstein distances.

Alain’s publications span prestigious journals like the Annals of Applied Probability and SIAM, with a focus on algorithmic innovation and rigorous mathematical foundations.