Conference « AI, Science and Society » of the Artificial Intelligence Action Summit Palaiseau, France, February 2025.. Stochastic Optimization beyond Gradient for Machine Learning. [Slides.pdf]
(plenary speaker) Monte Carlo and Quasi-Monte Carlo methods in Scientific Computing, Waterloo, Canada, August 2024. Stochastic Approximation beyond Gradient [slides.pdf] or [slides.pdf]
(plenary speaker) - French-German-Spanish conference on Optimization, Gijon, Spain, June 2024. Stochastic Approximation beyond Gradient [slides.pdf] or [slides.pdf]
European Meeting of Statisticians; Warsaw, Poland, July 2023. Stochastic Approximation beyond Gradient [slides.pdf] or [slides.pdf]
Conference « Processus markoviens, semi-markoviens et leurs applications »; Montpellier, France, June 2023. When Markov chains control Monte Carlo sampling. [slides.pdf] or [slides.pdf]
Conference « Learning and Optimization in Luminy », CIRM; Marseille, France, October 2022. Stochastic Variable Metric Forward-Backward with variance reduction for non-convex optimization [slides.pdf] or [slides.pdf]
Workshop « Current developments in MCMC methods », Banach Center; Warsaw, Poland; December 2021.. Federated Expectation Maximization with heterogeneity mitigation and variance reduction. [slides.pdf] or [slides.pdf]
Conference « Future Synergies for Stochastic and Learning algorithms », CIRM; Marseille, France; September 2021
Conference in Numerical Probability, in honour of Gilles Pagès; Paris, France; September 2020 (postponed in May 2021). A Variance Reduced Expectation Maximization algorithm for finite-sum optimization [slides.pdf and video] or [slides and video]
Conference « Structural Inference in High Dimensional Models »; Bordeaux, France; August 2020 (Canceled, Covid19)
Summer School 2020, Indo-French Centre for Applied Mathematics; Bangalore, India; July 2020 (Canceled, Covid19)
Conference « Optimization for Machine Learning », CIRM; Marseille, France; March 2020. Fast Incremental Expectation Maximization algorithm: how many iterations for an \epsilon-stationary point ? [slides.pdf] of [slides.pdf]
(plenary speaker) BayesComp 2020; Gainesville, USA; January 2020. Invitation declined for professional duties (new schedule of a HCERES evaluation).
(3 lectures and a talk) Advances in Applied Probability (ICTS Program); Bengaluru, India; August 2019. When Monte Carlo and Optimization met in a Markovian dance [abstract.pdf] or [abstract.pdf]
Lecture 1 : Computational Statistical Learning [slides.pdf, channel] or [slides, channel]
Lecture 2 : Controlled Markov chains [slides.pdf, channel] or [slides, channel]
Lecture 3 : Stochastic Approximation with Markovian Dynamics [slides.pdf, channel] or [slides, channel]
Talk: Stochastic Proximal-Gradient based algorithms: is Nesterov acceleration efficient ? [slides.pdf, channel] or [slides, channel]
(plenary speaker) Twelth International Conference on Monte Carlo methods and Applications (MCM2019); Sydney, Australie; July 2019. Monte Carlo methods and Optimization: Intertwinings [slides.pdf, group photo] or [slides, group photo]
Workshop « The mathematics of imaging »; Paris, France; February 2019. Stochastic Approximation-based algorithms, when the Monte Carlo bias does not vanish [slides.pdf, video] or [slides, video].
Workshop « Computational Statistics and Molecular Simulations: a practical cross-fertilization », BIRS; Oaxaca, Mexico; November 2018.. Convergence and Efficiency of Adaptive Importance Sampling techniques with partial biasing [slides.pdf, video] or [slides, video]
Workshop « Operator Splitting methods in Data Analysis »; Raleigh, USA; March 2018. Perturbed (accelerated) Proximal-Gradient Algorithms [slides.pdf] or [slides.pdf]
Foundations of Computational Mathematics (FOCM), Workshop « Stochastic Computation »; Barcelona, Spain; July 2017. Beyond Well-tempered Metadynamics algorithms for sampling multimodal target densities [slides.pdf] or [slides.pdf]
11th International Conference on Monte Carlo methods and Applications; Montreal, Canada; July 2017. MCMC design-based non-parametric regression for rare-event. Application to nested rosk computations [slides.pdf] or [slides.pdf]
International Conference on Monte Carlo techniques (closing conferene of a thematic cycle); Paris, France; July 2016. Nested risk computations through non parametric regression with Markovian design [slides.pdf] or [slides.pdf]
Workshop « Stochastic Algorithms for Big Data »; Paris, France; July 2016. Stochastic Perturbations of Proximal-Gradient methods for non-smooth convex optimization: the price of Markovian perturbations [slides.pdf] or [slides.pdf]
Workshop « High Dimensional Statistical Models & Big Data », Alan Turing Institute; London, United Kingdom; February 2016. Convergence of Perturbed Gradient-based methods for non-smooth convex optimization [slides.pdf] or [slides.pdf]
Workshop « Free energy calculations: a mathematical perspective », BIRS; Oaxaca, Mexico; July 2015. Mathematical aspects of adaptive samplers: application to free energy calculation. [slides.pdf] or [slides.pdf] and [video][photo]
International Conference « 7th Journées de Statistique du Sud »; Barcelona, Spain; June 2014. Sampling multimodal densities on large dimensional spaces [slides.pdf, photo] or [slides.pdf, photo]
Workshop « Computational methods for statistical mechanics »; Edinburgh, United Kingdom; July 2014. Convergence and Efficiency of the Wang Landau algorithm [slides.pdf] or [slides.pdf]
Workshop « From spectral gap to particle filters »; Reading, United Kingdom; September 2013. Adaptive and Interacting Markov chain Monte Carlo [slides.pdf] or [slides.pdf]
Workshop « New directions in Monte Carlo methods »; Gainesville, USA; January 2013. Convergence and Efficiency of the Wang Landau algorithm [slides.pdf, photo] or [slides, photo]
Workshop « Big Bang, Big Data, Big Computers »; Paris, France; September 2012. Adaptive abd Interacting Monte Carlo methods for bayesian analysis [slides.pdf] or [slides.pdf]
ISBA conference 2012; Kyoto, Japan; June 2012. Adaptive Equi-Energy samplers [slides.pdf] or [slides.pdf]
Workshop « Advances in Markov chain Monte Carlo »; Edinburgh, United Kingdom; April 2012. Stochastic Approximation-based adaptation for Interacting MCMC [slides.pdf] or [slides.pdf]
Workshop « Challenges and Advances in High Dimensional and High Complexity Monte Carlo Computation and Theory »; Calgary, Canada; March 2012. Parallel Tempering and interacting algorithms – Part II: adaptive equi-enery samplers [slides.pdf, photo] or [slides.pdf, photo]
Conference MCQMC 2010; Warsaw, Poland; August 2010. Convergence of Adaptive and Interacting MCMC algorithms [slides.pdf] or [slides.pdf]
Conference « Optimization in MCMC »; Warwick, United Kingdom; June 2009. Adaptive MCMC: theory and methods [slides.pdf] or [slides.pdf]
Congrès SSC-SFDS; Ottawa, Canada; May 2008. Stability of Markov chains based on fluid limit techniques. Applications to MCMC [slides.pdf] or [slides.pdf]
ADAP’ski; Bormio, Italy; January 2008. Fluid limit-based tuning of some hybrid MCMC samplers [slides.pdf] or [slides.pdf]
Workshop « New developments in MCMC: diffusions, images and other challenges »; Warwick, United Kingdom; August 2006. Criteria for subgeometric ergodicity of strong Markov processes [slides.pdf] or [slides.pdf]
Workshop « MCMC methodology »; Lancaster, United Kingdom; December 2001. Talk 1 « Some recent results on Hybrid samplers » [slides.ps] or [slides.ps]
Workshop « MCMC methodology »; Lancaster, United Kingdom; December 2001. Talk 2 « Convergence of the MCEM algorithm » [slides.ps] or [slides.ps]
European conference on Spatial and Computational Statistics; Ambleside, United Kingdom; September 2000.