Unbiased Markov Chain Monte Carlo: what, why, and how
Methodology
2024-06-12 v1
Abstract
This document presents methods to remove the initialization or burn-in bias from Markov chain Monte Carlo (MCMC) estimates, with consequences on parallel computing, convergence diagnostics and performance assessment. The document is written as an introduction to these methods for MCMC users. Some theoretical results are mentioned, but the focus is on the methodology.
Keywords
Cite
@article{arxiv.2406.06851,
title = {Unbiased Markov Chain Monte Carlo: what, why, and how},
author = {Yves F. Atchadé and Pierre E. Jacob},
journal= {arXiv preprint arXiv:2406.06851},
year = {2024}
}
Comments
To appear in the second edition of the handbook of MCMC