May 2022 Geometric convergence bounds for Markov chains in Wasserstein distance based on generalized drift and contraction conditions
Qian Qin, James P. Hobert
Author Affiliations +
Ann. Inst. H. Poincaré Probab. Statist. 58(2): 872-889 (May 2022). DOI: 10.1214/21-AIHP1195

Abstract

Let (Xn)n=0 denote a Markov chain on a Polish space that has a stationary distribution ϖ. This article concerns upper bounds on the Wasserstein distance between the distribution of Xn and ϖ. In particular, an explicit geometric bound on the distance to stationarity is derived using generalized drift and contraction conditions whose parameters vary across the state space. These new types of drift and contraction allow for sharper convergence bounds than the standard versions, whose parameters are constant. Application of the result is illustrated in the context of a non-linear autoregressive process and a Gibbs algorithm for a random effects model.

Soit (Xn)n=0 une chaîne de Markov définie sur un espace polonais qui a une distribution stationnaire ϖ. Cet article s’intéresse aux bornes supérieures pour la distance de Wasserstein entre les distributions Xn et ϖ. En particulier, une borne géométrique explicite est obtenue sur la distance à l’équilibre en utilisant des conditions de dérive et de contraction dont les paramètres varient dans l’espace d’états. Ces nouveaux types de dérive et contraction permettent d’obtenir des bornes de convergence plus précises que les versions standard où les paramètres sont constants. Des applications de ce résultat sont données dans le contexte des processus auto-régressifs non-linéaires et dans le contexte d’un algorithme de Gibbs pour le modèle à effets aléatoires.

Funding Statement

The second author was supported by NSF Grant DMS-15-11945.

Acknowledgements

We thank the Editor, the Associate Editor, and two anonymous reviewers for helpful comments and suggestions.

Citation

Download Citation

Qian Qin. James P. Hobert. "Geometric convergence bounds for Markov chains in Wasserstein distance based on generalized drift and contraction conditions." Ann. Inst. H. Poincaré Probab. Statist. 58 (2) 872 - 889, May 2022. https://doi.org/10.1214/21-AIHP1195

Information

Received: 6 September 2019; Revised: 15 February 2021; Accepted: 28 May 2021; Published: May 2022
First available in Project Euclid: 15 May 2022

MathSciNet: MR4421611
zbMATH: 1494.60077
Digital Object Identifier: 10.1214/21-AIHP1195

Subjects:
Primary: 60J05

Keywords: Convergence analysis , exponential convergence , Kantorovich–Rubinstein distance , Lyapunov drift function , Polish space , quantitative bound

Rights: Copyright © 2022 Association des Publications de l’Institut Henri Poincaré

JOURNAL ARTICLE
18 PAGES

This article is only available to subscribers.
It is not available for individual sale.
+ SAVE TO MY LIBRARY

Vol.58 • No. 2 • May 2022
Back to Top