Spectral gap of Metropolis-within-Gibbs under log-concavity

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Secchi, Cecilia, Zanella, Giacomo
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866918151629832192
author Secchi, Cecilia
Zanella, Giacomo
author_facet Secchi, Cecilia
Zanella, Giacomo
contents The Metropolis-within-Gibbs (MwG) algorithm is a widely used Markov Chain Monte Carlo method for sampling from high-dimensional distributions when exact conditional sampling is intractable. We study MwG with Random Walk Metropolis (RWM) updates, using proposal variances tuned to match the target's conditional variances. Assuming the target $π$ is a $d$-dimensional log-concave distribution with condition number $κ$, we establish a spectral gap lower bound of order $\mathcal{O}(1/κd)$ for the random-scan version of MwG, improving on the previously available $\mathcal{O}(1/κ^2 d)$ bound. This is obtained by developing sharp estimates of the conductance of one-dimensional RWM kernels, which can be of independent interest. The result shows that MwG can mix substantially faster with variance-adaptive proposals and that its mixing performance is just a constant factor worse than that of the exact Gibbs sampler, thus providing theoretical support to previously observed empirical behavior.
format Preprint
id arxiv_https___arxiv_org_abs_2509_26175
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Spectral gap of Metropolis-within-Gibbs under log-concavity
Secchi, Cecilia
Zanella, Giacomo
Machine Learning
Statistics Theory
Methodology
The Metropolis-within-Gibbs (MwG) algorithm is a widely used Markov Chain Monte Carlo method for sampling from high-dimensional distributions when exact conditional sampling is intractable. We study MwG with Random Walk Metropolis (RWM) updates, using proposal variances tuned to match the target's conditional variances. Assuming the target $π$ is a $d$-dimensional log-concave distribution with condition number $κ$, we establish a spectral gap lower bound of order $\mathcal{O}(1/κd)$ for the random-scan version of MwG, improving on the previously available $\mathcal{O}(1/κ^2 d)$ bound. This is obtained by developing sharp estimates of the conductance of one-dimensional RWM kernels, which can be of independent interest. The result shows that MwG can mix substantially faster with variance-adaptive proposals and that its mixing performance is just a constant factor worse than that of the exact Gibbs sampler, thus providing theoretical support to previously observed empirical behavior.
title Spectral gap of Metropolis-within-Gibbs under log-concavity
topic Machine Learning
Statistics Theory
Methodology
url https://arxiv.org/abs/2509.26175