APPLIED PROBABILITY AND STATISTICS SEMINAR

Fall 2006 / Spring 2007
Upcoming Talks / Past talks


Time:Wed Mar 21, 1.30pm-2.30pm

Place: PHYS 129

Speaker: Liangkun Li

Title: Markov Chain Monte Carlo and Gibbs Sampling

Abstract: The popular Bayesian appeoaches to obtain the posterior distribution requires the integration. However, when facing high-dimensional integration, the computation of Bayesian appoach is very difficult. Markov Chain Monte Carlo (MCMC) methods were proposed to solve such high-dimensional problems. MCMC is using the previous sample values to randomly generate the next sample variable to create a Markov Chain. Gibbs sampler is one particular MCMC method and widely used in those problems mentioned above.