Séance Séminaire

Séminaire de Probabilités et Statistique

Monte Carlo Sampling and Deep Generative Models for Bayesian Inference

Generating lifelike data with intricate patterns, like images, sound, or molecules, often requires powerful models. These models help us understand and approximate data patterns in complex spaces. However, even with advanced technology, building these models remains challenging. In this presentation, I will discuss recent advances to address this challenge by hybridising traditional sampling techniques with neural networks. I will discuss a stochastic plug-and-play (PnP) sampling algorithm that leverages variable splitting to efficiently sample from a posterior distribution. The algorithm based on split Gibbs sampling (SGS) divides the challenging task of posterior sampling into two simpler sampling problems. The first problem depends on the likelihood function, while the second is interpreted as a Bayesian denoising problem that can be readily carried out by a deep generative model, more specifically a diffusion model.

Séminaire en salle 109, également retransmis sur zoom : https://umontpellier-fr.zoom.us/j/94087408185