Geometric PDEs on weighted graphs for semi-supervised classification - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2014

Geometric PDEs on weighted graphs for semi-supervised classification

Résumé

In this paper, we consider the adaptation of two Partial Differential Equations (PDEs) on weighted graphs, p-Laplacian and eikonal equations, for semi-supervised classifi-cation tasks. These equations are a discrete analogue of well known geometric PDEs, which are widely used in image pro-cessing. While the p-Laplacian on graphs was intensively used in data classification, few works relate to the eikonal equation for data classification. The methods are illustrated through semi-supervised classification tasks on databases, where we compare the two algorithms. The results show that these methods perform well regarding the state-of-the-art and are applicable to the task of semi-supervised classification.
Fichier principal
Vignette du fichier
Toutain_ICMLA14.pdf (1.48 Mo) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-01108860 , version 1 (23-01-2015)

Identifiants

  • HAL Id : hal-01108860 , version 1

Citer

Matthieu Toutain, Abderrahim Elmoataz, Olivier Lézoray. Geometric PDEs on weighted graphs for semi-supervised classification. International Conference on Machine Learning and Applications, IEEE, 2012, Detroit, United States. ⟨hal-01108860⟩
127 Consultations
224 Téléchargements

Partager

Gmail Facebook X LinkedIn More