Non-asymptotic convergence bounds for Sinkhorn iterates and their gradients: a coupling approach

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Greco, Giacomo, Noble, Maxence, Conforti, Giovanni, Durmus, Alain
Format: Preprint
Published: 2023
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908573877927936
author Greco, Giacomo
Noble, Maxence
Conforti, Giovanni
Durmus, Alain
author_facet Greco, Giacomo
Noble, Maxence
Conforti, Giovanni
Durmus, Alain
contents Computational optimal transport (OT) has recently emerged as a powerful framework with applications in various fields. In this paper we focus on a relaxation of the original OT problem, the entropic OT problem, which allows to implement efficient and practical algorithmic solutions, even in high dimensional settings. This formulation, also known as the Schrödinger Bridge problem, notably connects with Stochastic Optimal Control (SOC) and can be solved with the popular Sinkhorn algorithm. In the case of discrete-state spaces, this algorithm is known to have exponential convergence; however, achieving a similar rate of convergence in a more general setting is still an active area of research. In this work, we analyze the convergence of the Sinkhorn algorithm for probability measures defined on the $d$-dimensional torus $\mathbb{T}_L^d$, that admit densities with respect to the Haar measure of $\mathbb{T}_L^d$. In particular, we prove pointwise exponential convergence of Sinkhorn iterates and their gradient. Our proof relies on the connection between these iterates and the evolution along the Hamilton-Jacobi-Bellman equations of value functions obtained from SOC-problems. Our approach is novel in that it is purely probabilistic and relies on coupling by reflection techniques for controlled diffusions on the torus.
format Preprint
id arxiv_https___arxiv_org_abs_2304_06549
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Non-asymptotic convergence bounds for Sinkhorn iterates and their gradients: a coupling approach
Greco, Giacomo
Noble, Maxence
Conforti, Giovanni
Durmus, Alain
Probability
Optimization and Control
Machine Learning
49Q22, 93E20 (Primary) 49N05, 90C25, 47D07
Computational optimal transport (OT) has recently emerged as a powerful framework with applications in various fields. In this paper we focus on a relaxation of the original OT problem, the entropic OT problem, which allows to implement efficient and practical algorithmic solutions, even in high dimensional settings. This formulation, also known as the Schrödinger Bridge problem, notably connects with Stochastic Optimal Control (SOC) and can be solved with the popular Sinkhorn algorithm. In the case of discrete-state spaces, this algorithm is known to have exponential convergence; however, achieving a similar rate of convergence in a more general setting is still an active area of research. In this work, we analyze the convergence of the Sinkhorn algorithm for probability measures defined on the $d$-dimensional torus $\mathbb{T}_L^d$, that admit densities with respect to the Haar measure of $\mathbb{T}_L^d$. In particular, we prove pointwise exponential convergence of Sinkhorn iterates and their gradient. Our proof relies on the connection between these iterates and the evolution along the Hamilton-Jacobi-Bellman equations of value functions obtained from SOC-problems. Our approach is novel in that it is purely probabilistic and relies on coupling by reflection techniques for controlled diffusions on the torus.
title Non-asymptotic convergence bounds for Sinkhorn iterates and their gradients: a coupling approach
topic Probability
Optimization and Control
Machine Learning
49Q22, 93E20 (Primary) 49N05, 90C25, 47D07
url https://arxiv.org/abs/2304.06549