Speeeding Up Markov Chain Monte Carlo Algorithms

We prove an upper bound on the convergence rate of Markov Chain Monte Carlo (MCMC) algorithms for the important special case when the state space can be aggregated into a smaller space, such that the aggregated chain approximately preserves the Markov property.

• Type: Conference
• Year: 2006
• Where: FCS
• Authors: Andras Farago
• Keywords: FCS 2006, FCS 2007, Important Special Case, Markov Chain Monte Carlo, Markov Property