Using Sinkhorn in the JKO Scheme Adds Linear Diffusion

Abstract. The JKO (Jordan–Kinderlehrer–Otto) scheme is a time-discrete scheme of implicit Euler type that allows one to construct weak solutions of evolution PDEs which have a Wasserstein gradient structure. The purpose of this work is to study the effect of replacing the classical quadratic optimal transport problem by the Schrödinger problem (a.k.a. the entropic regularization of optimal transport, efficiently computed by the Sinkhorn algorithm) at each step of this scheme. We find that if [Formula: see text] is the regularization parameter of the Schrödinger problem, and [Formula: see text] is the time-step parameter, considering the limit [Formula: see text] with [Formula: see text] results in adding the term [Formula: see text] on the right-hand side of the limiting PDE. In the case [Formula: see text] we improve a previous result by Carlier et al. [ SIAM J. Math. Anal., 49 (2017), pp. 1385–1418].

Authors

Institutions

Publication Details

Journal
SIAM Journal on Mathematical Analysis
Published
2026-07-27
DOI
https://doi.org/10.1137/25m1736001
Primary Topic
Seismic Imaging and Inversion Techniques
Type
article
Field-Weighted Citation Impact
0.00

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

Using Sinkhorn in the JKO Scheme Adds Linear Diffusion

Filippo Santambrogio, Aymeric Baradat, Anastasiia Hraivoronska
SIAM Journal on Mathematical Analysis
Seismic Imaging and Inversion Techniques
article

Using Sinkhorn in the JKO Scheme Adds Linear Diffusion

Filippo Santambrogio, Aymeric Baradat, Anastasiia Hraivoronska
article en

Abstract

Abstract. The JKO (Jordan–Kinderlehrer–Otto) scheme is a time-discrete scheme of implicit Euler type that allows one to construct weak solutions of evolution PDEs which have a Wasserstein gradient structure. The purpose of this work is to study the effect of replacing the classical quadratic optimal transport problem by the Schrödinger problem (a.k.a. the entropic regularization of optimal transport, efficiently computed by the Sinkhorn algorithm) at each step of this scheme. We find that if [Formula: see text] is the regularization parameter of the Schrödinger problem, and [Formula: see text] is the time-step parameter, considering the limit [Formula: see text] with [Formula: see text] results in adding the term [Formula: see text] on the right-hand side of the limiting PDE. In the case [Formula: see text] we improve a previous result by Carlier et al. [ SIAM J. Math. Anal., 49 (2017), pp. 1385–1418].

SIAM Journal on Mathematical AnalysisVol. 58(4)
Université Claude Bernard Lyon 1 (FR), Centre National de la Recherche Scientifique (FR), Institut Camille Jordan (FR), Institut National des Sciences Appliquées de Lyon (FR), Université Jean Monnet (FR)
H2020 European Research Council
Openalex Percentile: Top 100%
Seismic Imaging and Inversion Techniques
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.