Unsupervised collaborative learning based on Optimal Transport theory

Collaborative learning has recently achieved very significant results. It still suffers, however, from several issues, including the type of information that needs to be exchanged, the criteria for stopping and how to choose the right collaborators. We aim in this paper to improve the quality of the...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Ben-Bouazza Fatima-Ezzahraa, Bennani Younès, Cabanes Guénaël, Touzani Abdelfettah
Formato:	article
Lenguaje:	EN
Publicado:	De Gruyter 2021
Materias:	collaborative learning distributed learning optimal transport sinkhorn matrix wasserstein distance Science Q Electronic computers. Computer science QA75.5-76.95
Acceso en línea:	https://doaj.org/article/31948ad869ff4e72a682728e318471c6
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:31948ad869ff4e72a682728e318471c6
record_format	dspace
spelling	oai:doaj.org-article:31948ad869ff4e72a682728e318471c62021-12-05T14:10:51ZUnsupervised collaborative learning based on Optimal Transport theory2191-026X10.1515/jisys-2020-0068https://doaj.org/article/31948ad869ff4e72a682728e318471c62021-06-01T00:00:00Zhttps://doi.org/10.1515/jisys-2020-0068https://doaj.org/toc/2191-026XCollaborative learning has recently achieved very significant results. It still suffers, however, from several issues, including the type of information that needs to be exchanged, the criteria for stopping and how to choose the right collaborators. We aim in this paper to improve the quality of the collaboration and to resolve these issues via a novel approach inspired by Optimal Transport theory. More specifically, the objective function for the exchange of information is based on the Wasserstein distance, with a bidirectional transport of information between collaborators. This formulation allows to learns a stopping criterion and provide a criterion to choose the best collaborators. Extensive experiments are conducted on multiple data-sets to evaluate the proposed approach.Ben-Bouazza Fatima-EzzahraaBennani YounèsCabanes GuénaëlTouzani AbdelfettahDe Gruyterarticlecollaborative learningdistributed learningoptimal transportsinkhorn matrixwasserstein distanceScienceQElectronic computers. Computer scienceQA75.5-76.95ENJournal of Intelligent Systems, Vol 30, Iss 1, Pp 698-719 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	collaborative learning distributed learning optimal transport sinkhorn matrix wasserstein distance Science Q Electronic computers. Computer science QA75.5-76.95
spellingShingle	collaborative learning distributed learning optimal transport sinkhorn matrix wasserstein distance Science Q Electronic computers. Computer science QA75.5-76.95 Ben-Bouazza Fatima-Ezzahraa Bennani Younès Cabanes Guénaël Touzani Abdelfettah Unsupervised collaborative learning based on Optimal Transport theory
description	Collaborative learning has recently achieved very significant results. It still suffers, however, from several issues, including the type of information that needs to be exchanged, the criteria for stopping and how to choose the right collaborators. We aim in this paper to improve the quality of the collaboration and to resolve these issues via a novel approach inspired by Optimal Transport theory. More specifically, the objective function for the exchange of information is based on the Wasserstein distance, with a bidirectional transport of information between collaborators. This formulation allows to learns a stopping criterion and provide a criterion to choose the best collaborators. Extensive experiments are conducted on multiple data-sets to evaluate the proposed approach.
format	article
author	Ben-Bouazza Fatima-Ezzahraa Bennani Younès Cabanes Guénaël Touzani Abdelfettah
author_facet	Ben-Bouazza Fatima-Ezzahraa Bennani Younès Cabanes Guénaël Touzani Abdelfettah
author_sort	Ben-Bouazza Fatima-Ezzahraa
title	Unsupervised collaborative learning based on Optimal Transport theory
title_short	Unsupervised collaborative learning based on Optimal Transport theory
title_full	Unsupervised collaborative learning based on Optimal Transport theory
title_fullStr	Unsupervised collaborative learning based on Optimal Transport theory
title_full_unstemmed	Unsupervised collaborative learning based on Optimal Transport theory
title_sort	unsupervised collaborative learning based on optimal transport theory
publisher	De Gruyter
publishDate	2021
url	https://doaj.org/article/31948ad869ff4e72a682728e318471c6
work_keys_str_mv	AT benbouazzafatimaezzahraa unsupervisedcollaborativelearningbasedonoptimaltransporttheory AT bennaniyounes unsupervisedcollaborativelearningbasedonoptimaltransporttheory AT cabanesguenael unsupervisedcollaborativelearningbasedonoptimaltransporttheory AT touzaniabdelfettah unsupervisedcollaborativelearningbasedonoptimaltransporttheory
_version_	1718371663514435584

Unsupervised collaborative learning based on Optimal Transport theory

Ejemplares similares