Optimal Transport with Adaptive Regularisation - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2023

Optimal Transport with Adaptive Regularisation

Résumé

Regularising the primal formulation of optimal transport (OT) with a strictly convex term leads to enhanced numerical complexity and a denser transport plan. Many formulations impose a global constraint on the transport plan, for instance by relying on entropic regularisation. As it is more expensive to diffuse mass for outlier points compared to central ones, this typically results in a significant imbalance in the way mass is spread across the points. This can be detrimental for some applications where a minimum of smoothing is required per point. To remedy this, we introduce OT with Adaptive RegularIsation (OTARI), a new formulation of OT that imposes constraints on the mass going in or/and out of each point. We then showcase the benefits of this approach for domain adaptation.
Fichier principal
Vignette du fichier
workshop_paper.pdf (1.39 Mo) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-04229636 , version 1 (05-10-2023)

Identifiants

Citer

Hugues Van Assel, Titouan Vayer, Remi Flamary, Nicolas Courty. Optimal Transport with Adaptive Regularisation. NeurIPS OTML Workshop, Dec 2023, New Orleans, France. ⟨hal-04229636⟩
82 Consultations
152 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More