|TITLE:||Adaptive MCMC For Everyone|
|SPEAKER:||Jeffrey Rosenthal (University of Toronto)|
|DATE:||Vendredi, 1 Mai 2015|
Markov chain Monte Carlo (MCMC) algorithms, such as the Metropolis Algorithm and the Gibbs Sampler, are an extremely useful and popular method of approximately sampling from complicated probability distributions. Adaptive MCMC attempts to automatically modify the algorithm while it runs, to improve its performance on the fly. However, such adaption often destroys the ergodicity properties necessary for the algorithm to be valid. In this talk, we first illustrate basic MCMC algorithms using simple Java applets. We then discuss adaptive MCMC, and present results and examples concerning its ergodicity and efficiency. We close with some recent ideas which make adaptive MCMC more widely applicable for all users in broader contexts.
About the speaker: Jeffrey Rosenthal is a professor in the University of Toronto's Department of Statistics, cross-appointed with Department of Mathematics. He received his B.Sc. (in mathematics, physics, and computer science) from the University of Toronto in 1988, and his Ph.D. in mathematics from Harvard University in 1992 (supervised by Persi Diaconis). He is a recipient of the CRM-SSC Prize in 2006, the COPSS Presidents' Award in 2007 and the Statistical Society of Canada Gold Medal in 2013. He was elected a Fellow of the Royal Society of Canada in 2012.