Ergodicity of Langevin Dynamics and its Discretizations for Non-smooth Potentials

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Fruehwirth, Lorenz, Habring, Andreas
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916039014481920
author Fruehwirth, Lorenz
Habring, Andreas
author_facet Fruehwirth, Lorenz
Habring, Andreas
contents This article is concerned with sampling from Gibbs distributions $π(x)\propto e^{-U(x)}$ using Markov chain Monte Carlo methods. In particular, we investigate Langevin dynamics in the continuous- and the discrete-time setting for such distributions with potentials $U(x)$ which are strongly-convex but possibly non-differentiable. We show that the corresponding subgradient Langevin dynamics are exponentially ergodic to the target density $π$ in the continuous setting and that certain explicit as well as semi-implicit discretizations are geometrically ergodic and approximate $π$ for vanishing discretization step size. Moreover, we prove that the discrete schemes satisfy the law of large numbers allowing to use consecutive iterates of a Markov chain in order to compute statistics of the stationary distribution posing a significant reduction of computational complexity in practice. Numerical experiments are provided confirming the theoretical findings and showcasing the practical relevance of the proposed methods in imaging applications.
format Preprint
id arxiv_https___arxiv_org_abs_2411_12051
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Ergodicity of Langevin Dynamics and its Discretizations for Non-smooth Potentials
Fruehwirth, Lorenz
Habring, Andreas
Numerical Analysis
Optimization and Control
60J20, 68U10, 94A08
G.3; I.4.5
This article is concerned with sampling from Gibbs distributions $π(x)\propto e^{-U(x)}$ using Markov chain Monte Carlo methods. In particular, we investigate Langevin dynamics in the continuous- and the discrete-time setting for such distributions with potentials $U(x)$ which are strongly-convex but possibly non-differentiable. We show that the corresponding subgradient Langevin dynamics are exponentially ergodic to the target density $π$ in the continuous setting and that certain explicit as well as semi-implicit discretizations are geometrically ergodic and approximate $π$ for vanishing discretization step size. Moreover, we prove that the discrete schemes satisfy the law of large numbers allowing to use consecutive iterates of a Markov chain in order to compute statistics of the stationary distribution posing a significant reduction of computational complexity in practice. Numerical experiments are provided confirming the theoretical findings and showcasing the practical relevance of the proposed methods in imaging applications.
title Ergodicity of Langevin Dynamics and its Discretizations for Non-smooth Potentials
topic Numerical Analysis
Optimization and Control
60J20, 68U10, 94A08
G.3; I.4.5
url https://arxiv.org/abs/2411.12051