A seminar by Minh-Ngoc Tran from University of Sydney
Title: Particle Variational Bayes
Abstract: Variational Bayes (VB) is widely recognised as a highly efficient and scalable technique for Bayesian inference. However, classical VB often imposes restrictions on the space of variational distributions, typically restricting it to a specific set of parametric distributions or factorized distributions. In this talk, we explore ways to relax these restrictions by traversing a set of particles to approximate the target distribution. The theoretical basis of the new particle VB method can be established using Optimal Transport theory, which allows us to make the space of probability measures into a differential manifold.
This talk focuses particularly on the novel Particle Mean Field Variational Bayes (PMFVB) approach, which extends the classical MFVB method without requiring conjugate priors or analytical calculations. We also discuss the connection between the optimisation-based Particle VB and sampling-based Sequential Monte Carlo, and how to improve the latter. We demonstrate the effectiveness of the PMFVB approach using Bayesian logistic regression, stochastic volatility, and deep neural networks.
For further information, please contact RSFAS Seminars.