UC Irvine Today
Statistics Seminar Series: Jumping Between Modes - A New Markov Chain Monte Carlo Sampler: Yaming Yu
The Statistics Seminar Series is proud to present Yaming Yu, Associate Professor, UCI. A multi-modal posterior distribution is a common difficulty in Bayesian computation. We propose a new Markov chain Monte Carlo method to address this problem. This method runs multiple chains in the parallel, with occasional jumps from one chain to the neighborhood of other chains. Unlike parallel-tempering or many related population-based MCMC methods, however, there is no temperature parameter and the same target density is maintained for each chain.
Thursday, November 5 at 4:00pm to 5:00pmVirtual Event