Cargando…

Efficient Discretization of Optimal Transport

Obtaining solutions to optimal transportation (OT) problems is typically intractable when marginal spaces are continuous. Recent research has focused on approximating continuous solutions with discretization methods based on i.i.d. sampling, and this has shown convergence as the sample size increase...

Descripción completa

Detalles Bibliográficos
Autores principales: Wang, Junqi, Wang, Pei, Shafto, Patrick
Formato: Online Artículo Texto
Lenguaje:English
Publicado: MDPI 2023
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10297305/
https://www.ncbi.nlm.nih.gov/pubmed/37372183
http://dx.doi.org/10.3390/e25060839
_version_ 1785063852334907392
author Wang, Junqi
Wang, Pei
Shafto, Patrick
author_facet Wang, Junqi
Wang, Pei
Shafto, Patrick
author_sort Wang, Junqi
collection PubMed
description Obtaining solutions to optimal transportation (OT) problems is typically intractable when marginal spaces are continuous. Recent research has focused on approximating continuous solutions with discretization methods based on i.i.d. sampling, and this has shown convergence as the sample size increases. However, obtaining OT solutions with large sample sizes requires intensive computation effort, which can be prohibitive in practice. In this paper, we propose an algorithm for calculating discretizations with a given number of weighted points for marginal distributions by minimizing the (entropy-regularized) Wasserstein distance and providing bounds on the performance. The results suggest that our plans are comparable to those obtained with much larger numbers of i.i.d. samples and are more efficient than existing alternatives. Moreover, we propose a local, parallelizable version of such discretizations for applications, which we demonstrate by approximating adorable images.
format Online
Article
Text
id pubmed-10297305
institution National Center for Biotechnology Information
language English
publishDate 2023
publisher MDPI
record_format MEDLINE/PubMed
spelling pubmed-102973052023-06-28 Efficient Discretization of Optimal Transport Wang, Junqi Wang, Pei Shafto, Patrick Entropy (Basel) Article Obtaining solutions to optimal transportation (OT) problems is typically intractable when marginal spaces are continuous. Recent research has focused on approximating continuous solutions with discretization methods based on i.i.d. sampling, and this has shown convergence as the sample size increases. However, obtaining OT solutions with large sample sizes requires intensive computation effort, which can be prohibitive in practice. In this paper, we propose an algorithm for calculating discretizations with a given number of weighted points for marginal distributions by minimizing the (entropy-regularized) Wasserstein distance and providing bounds on the performance. The results suggest that our plans are comparable to those obtained with much larger numbers of i.i.d. samples and are more efficient than existing alternatives. Moreover, we propose a local, parallelizable version of such discretizations for applications, which we demonstrate by approximating adorable images. MDPI 2023-05-24 /pmc/articles/PMC10297305/ /pubmed/37372183 http://dx.doi.org/10.3390/e25060839 Text en © 2023 by the authors. https://creativecommons.org/licenses/by/4.0/Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
spellingShingle Article
Wang, Junqi
Wang, Pei
Shafto, Patrick
Efficient Discretization of Optimal Transport
title Efficient Discretization of Optimal Transport
title_full Efficient Discretization of Optimal Transport
title_fullStr Efficient Discretization of Optimal Transport
title_full_unstemmed Efficient Discretization of Optimal Transport
title_short Efficient Discretization of Optimal Transport
title_sort efficient discretization of optimal transport
topic Article
url https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10297305/
https://www.ncbi.nlm.nih.gov/pubmed/37372183
http://dx.doi.org/10.3390/e25060839
work_keys_str_mv AT wangjunqi efficientdiscretizationofoptimaltransport
AT wangpei efficientdiscretizationofoptimaltransport
AT shaftopatrick efficientdiscretizationofoptimaltransport