Generalized Wasserstein Barycenters

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Tornabene, Francesco, Veneroni, Marco, Savaré, Giuseppe
Format: Preprint
Veröffentlicht: 2024
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866912463743614976
author Tornabene, Francesco
Veneroni, Marco
Savaré, Giuseppe
author_facet Tornabene, Francesco
Veneroni, Marco
Savaré, Giuseppe
contents We study the existence and uniqueness of the barycenter of a signed distribution of probability measures on a Hilbert space. The barycenter is found, as usual, as a minimum of a functional. In the case where the positive part of the signed measure is atomic, we can show also uniqueness. In the one-dimensional case, we characterize the quantile function of the unique minimum as the orthogonal projection of the $L^2$-barycenter of the quantiles on the cone of nonincreasing functions in $L^2(0,1)$. Further, we provide a stability estimate in dimension one and a counterexample to uniqueness in $\mathbb{R}^2$. Finally, we address the consistency of the barycenters and we prove that barycenters of a sequence of approximating measures converge (up to subsequences) to a barycenter of the limit measure.
format Preprint
id arxiv_https___arxiv_org_abs_2411_06838
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Generalized Wasserstein Barycenters
Tornabene, Francesco
Veneroni, Marco
Savaré, Giuseppe
Probability
Functional Analysis
Optimization and Control
49J27, 49J45, 49Q22
We study the existence and uniqueness of the barycenter of a signed distribution of probability measures on a Hilbert space. The barycenter is found, as usual, as a minimum of a functional. In the case where the positive part of the signed measure is atomic, we can show also uniqueness. In the one-dimensional case, we characterize the quantile function of the unique minimum as the orthogonal projection of the $L^2$-barycenter of the quantiles on the cone of nonincreasing functions in $L^2(0,1)$. Further, we provide a stability estimate in dimension one and a counterexample to uniqueness in $\mathbb{R}^2$. Finally, we address the consistency of the barycenters and we prove that barycenters of a sequence of approximating measures converge (up to subsequences) to a barycenter of the limit measure.
title Generalized Wasserstein Barycenters
topic Probability
Functional Analysis
Optimization and Control
49J27, 49J45, 49Q22
url https://arxiv.org/abs/2411.06838