REMINDER! Royal Statistical Society Statistical Computing Section Meeting Wednesday 20 January, 2.00 - 5.00pm at the RSS (Tea 3:30pm) Gibbs sampling, WinBUGS and perfect simulation ---------------------------------------------- Speakers KAREN VINES (Open University) DAVID SPIEGELHALTER (MRC Biostatistics Unit Cambridge) ANDREW THOMAS (Imperial College London) PETER GREEN (University of Bristol) Programme --------- 2.00pm Karen Vines Introduction to MCMC and graphical models 2.45pm David Spiegelhalter & WinBUGS Andrew Thomas 3.30pm Tea 4.00pm Peter Green Perfect simulation, and its possibilities in Bayesian statistics Synopses -------- KAREN VINES (Open University) Introduction to MCMC and graphical models This talk aims to provide a basic overview of Markov chain Monte Carlo (MCMC) techniques for the novice practitioner. Starting from the use of graphical models to identify exploitable conditional independence relationships and going through to interpretation of the output, we will cover issues that can arise in any analysis. DAVID SPIEGELHALTER (MRC Biostatistics Unit Cambridge) and ANDREW THOMAS (Imperial College London) WinBUGS WinBUGS performs MCMC analysis on Bayesian graphical models. We shall discuss the philosophy behind the program and demonstrate some recent developments. WinBUGS is freely available from http://www.mrc-bsu.cam.ac.uk/bugs. PETER GREEN (University of Bristol) Perfect simulation, and its possibilities in Bayesian statistics With the ultimate objective of Bayesian MCMC with guaranteed convergence, the purpose of this talk is to describe recent efforts to construct exact sampling methods for continuous-state Markov chains. We will consider whether application of such methods could become sufficiently convenient that they could become the basis for routine Bayesian computation in the foreseeable future. This is joint work with Duncan Murdoch. ---------------------- Suzanne Evans [log in to unmask] %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%