Translator Disclaimer
August 2020 On sampling from a log-concave density using kinetic Langevin diffusions
Arnak S. Dalalyan, Lionel Riou-Durand
Bernoulli 26(3): 1956-1988 (August 2020). DOI: 10.3150/19-BEJ1178

Abstract

Langevin diffusion processes and their discretizations are often used for sampling from a target density. The most convenient framework for assessing the quality of such a sampling scheme corresponds to smooth and strongly log-concave densities defined on $\mathbb{R}^{p}$. The present work focuses on this framework and studies the behavior of the Monte Carlo algorithm based on discretizations of the kinetic Langevin diffusion. We first prove the geometric mixing property of the kinetic Langevin diffusion with a mixing rate that is optimal in terms of its dependence on the condition number. We then use this result for obtaining improved guarantees of sampling using the kinetic Langevin Monte Carlo method, when the quality of sampling is measured by the Wasserstein distance. We also consider the situation where the Hessian of the log-density of the target distribution is Lipschitz-continuous. In this case, we introduce a new discretization of the kinetic Langevin diffusion and prove that this leads to a substantial improvement of the upper bound on the sampling error measured in Wasserstein distance.

Citation

Download Citation

Arnak S. Dalalyan. Lionel Riou-Durand. "On sampling from a log-concave density using kinetic Langevin diffusions." Bernoulli 26 (3) 1956 - 1988, August 2020. https://doi.org/10.3150/19-BEJ1178

Information

Received: 1 November 2018; Revised: 1 October 2019; Published: August 2020
First available in Project Euclid: 27 April 2020

zbMATH: 07193949
MathSciNet: MR4091098
Digital Object Identifier: 10.3150/19-BEJ1178

Rights: Copyright © 2020 Bernoulli Society for Mathematical Statistics and Probability

JOURNAL ARTICLE
33 PAGES

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

SHARE
Vol.26 • No. 3 • August 2020
Back to Top