Previous: Wednesday 16th June 2004 — Introduction to learning theory
Next: 28th June 2004 — Alan Yuille
21st June 2004 — Regeneration
Iain and Ed will present Regeneration a method for splitting Markov
chain Monte Carlo (MCMC) runs into independent tours. Originally designed as a
convergence analysis tool, Regeneration can also be used to dynamically improve
MCMC algorithms.
Adaptive Markov Chain
Monte Carlo through Regeneration, Walter R. Gilks, Gareth O. Roberts and
Sujit K. Sahu. 1998. The journal version is here
if you have access to Jstor. Journal of the American Statistical Association,
93, 1045--1054.
Iain Murray.