Resampling from the past to improve on MCMC algorithms
统计理论
2007-06-13 v1 概率论
统计理论
摘要
We introduce the idea that resampling from past observations in a Markov Chain Monte Carlo sampler can fasten convergence. We prove that proper resampling from the past does not disturb the limit distribution of the algorithm. We illustrate the method with two examples. The first on a Bayesian analysis of stochastic volatility models and the other on Bayesian phylogeny reconstruction.
引用
@article{arxiv.math/0605452,
title = {Resampling from the past to improve on MCMC algorithms},
author = {Yves F. Atchade},
journal= {arXiv preprint arXiv:math/0605452},
year = {2007}
}
备注
26 pages, 7 figures