Wasserstein Gradient Flows of the Discrepancy with Distance Kernel on the Line

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Hertrich, Johannes, Beinert, Robert, Gräf, Manuel, Steidl, Gabriele
Format: Preprint
Veröffentlicht: 2023
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866910561909866496
author Hertrich, Johannes
Beinert, Robert
Gräf, Manuel
Steidl, Gabriele
author_facet Hertrich, Johannes
Beinert, Robert
Gräf, Manuel
Steidl, Gabriele
contents This paper provides results on Wasserstein gradient flows between measures on the real line. Utilizing the isometric embedding of the Wasserstein space $\mathcal P_2(\mathbb R)$ into the Hilbert space $L_2((0,1))$, Wasserstein gradient flows of functionals on $\mathcal P_2(\mathbb R)$ can be characterized as subgradient flows of associated functionals on $L_2((0,1))$. For the maximum mean discrepancy functional $\mathcal F_ν:= \mathcal D^2_K(\cdot, ν)$ with the non-smooth negative distance kernel $K(x,y) = -|x-y|$, we deduce a formula for the associated functional. This functional appears to be convex, and we show that $\mathcal F_ν$ is convex along (generalized) geodesics. For the Dirac measure $ν= δ_q$, $q \in \mathbb R$ as end point of the flow, this enables us to determine the Wasserstein gradient flows analytically. Various examples of Wasserstein gradient flows are given for illustration.
format Preprint
id arxiv_https___arxiv_org_abs_2301_04441
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Wasserstein Gradient Flows of the Discrepancy with Distance Kernel on the Line
Hertrich, Johannes
Beinert, Robert
Gräf, Manuel
Steidl, Gabriele
Optimization and Control
Numerical Analysis
Probability
This paper provides results on Wasserstein gradient flows between measures on the real line. Utilizing the isometric embedding of the Wasserstein space $\mathcal P_2(\mathbb R)$ into the Hilbert space $L_2((0,1))$, Wasserstein gradient flows of functionals on $\mathcal P_2(\mathbb R)$ can be characterized as subgradient flows of associated functionals on $L_2((0,1))$. For the maximum mean discrepancy functional $\mathcal F_ν:= \mathcal D^2_K(\cdot, ν)$ with the non-smooth negative distance kernel $K(x,y) = -|x-y|$, we deduce a formula for the associated functional. This functional appears to be convex, and we show that $\mathcal F_ν$ is convex along (generalized) geodesics. For the Dirac measure $ν= δ_q$, $q \in \mathbb R$ as end point of the flow, this enables us to determine the Wasserstein gradient flows analytically. Various examples of Wasserstein gradient flows are given for illustration.
title Wasserstein Gradient Flows of the Discrepancy with Distance Kernel on the Line
topic Optimization and Control
Numerical Analysis
Probability
url https://arxiv.org/abs/2301.04441