Regeneration in Markov Chain
نویسنده
چکیده
Markov chain sampling has recently received considerable attention in particular in the context of Bayesian computation and maximum likelihood estimation. This paper discusses the use of Markov chain splitting, originally developed for the theoretical analysis of general state space Markov chains, to introduce regeneration into Markov chain samplers. This allows the use of regenerative methods for analyzing the output of these samplers, and can provide a useful diagnostic of sampler performance. The approach is applied to several samplers, including certain Metropolis samplers that can be used on their own or in hybrid samplers, and is illustrated in several examples.
منابع مشابه
Practical Regeneration for Markov Chain Monte Carlo Simulation
Regeneration is a useful tool in Markov chain Monte Carlo simulation, since it can be used to side-step the burn-in problem and to construct estimates of the variance of parameter estimates themselves. Unfortunately, it is often diÆcult to take advantage of, since for most chains, no recurrent atom exists, and it is not always easy to use Nummelin's splitting method to identify regeneration poi...
متن کاملAN APPLICATION OF TRAJECTORIES AMBIGUITY IN TWO-STATE MARKOV CHAIN
In this paper, the ambiguity of nite state irreducible Markov chain trajectories is reminded and is obtained for two state Markov chain. I give an applicable example of this concept in President election
متن کاملIdentification of Regeneration Times in MCMC Simulation, with Application to Adaptive Schemes
Regeneration is a useful tool in Markov chain Monte Carlo simulation because it can be used to side-step the burn-in problem and to construct better estimates of the variance of parameter estimates themselves. It also provides a simple way to introduce adaptive behavior into a Markov chain, and to use parallel processors to build a single chain. Regeneration is often difficult to take advantage...
متن کاملExact Simulation and Error-Controlled Sampling via Regeneration and a Bernoulli Factory
Methods using regeneration have been used to draw approximations to the stationary distribution of Markov Chain Monte Carlo processes. We introduce an algorithm that allows exact sampling of the stationary distribution through the use of a regeneration method and a Bernoulli Factory to select samples within each regeneration cycle that are shown to be from the desired density. We demonstrate th...
متن کاملAdaptive Chains
Adaptive chains are chains that are able to learn from all previous elements in the chain. It is an extension of Markov chains. It is proved convergence of adaptive chains that satisses a strong Doeblin condition (i.e., the transition density r from x i to y satisses r(y; x 1 ; x 2 ; : : : ; x i) a i (y) for all x 1 ; : : : ; x i ; y in the state space.) By using the previous iterations of the ...
متن کامل