Continuous-time MCMC, Paul Fearnhead (CSSS Seminar, 3/6/2019)
Posted: 3/4/2019 (Local Events)
Professor, Departments of Statistics, Lancaster University, https://www.maths.lancs.ac.uk/fearnhea/
Recently, there have been conceptually novel developments in Monte Carlo methods through the introduction of new MCMC algorithms which are based on continuous-time, rather than discrete-time, Markov processes. These show promise for scalable Bayesian Analysis: they naturally have non-reversible dynamics which enable them to mix faster in high-dimensional settings; sometimes they can be implemented in a way that requires access to only a small number of data points at each iteration, and yet still sample from the true posterior; and they automatically take account of sparsity in the dependence structure. This talk will give an overview of the recent work in this area.
Time: 12:30–1:30 PM
Location: Savery (SAV) 409