Classification of high dimensional data: High Dimensional Discriminant Analysis

Charles Bouveyron 1, 2 Stephane Girard 2 Cordelia Schmid 1, *
* Auteur correspondant
1 LEAR - Learning and recognition in vision
GRAVIR - IMAG - Graphisme, Vision et Robotique, Inria Grenoble - Rhône-Alpes, CNRS - Centre National de la Recherche Scientifique : FR71
Abstract : We propose a new method of discriminant analysis, called High Dimensional Discriminant Analysis (HHDA). Our approach is based on the assumption that high dimensional data live in dierent subspaces with low dimensionality. Thus, HDDA reduces the dimension for each class independently and regularizes class conditional covariance matrices in order to adapt the Gaussian framework to high dimensional data. This regularization is achieved by assuming that classes are spherical in their eigenspace. HDDA is applied to recognize objects in real images and its performances are compared to classical classication methods.
Type de document :
Communication dans un congrès
Subspace, Latent Structure and Feature Selection techniques: Statistical and Optimisation perspectives Workshop, Feb 2005, Bohinj, Slovenia. 2005
Liste complète des métadonnées

Littérature citée [5 références]  Voir  Masquer  Télécharger

https://hal.inria.fr/inria-00548517
Contributeur : Thoth Team <>
Soumis le : lundi 20 décembre 2010 - 09:08:57
Dernière modification le : jeudi 11 janvier 2018 - 06:20:04
Document(s) archivé(s) le : lundi 21 mars 2011 - 03:08:05

Fichier

BGS05a.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

  • HAL Id : inria-00548517, version 1

Collections

IMAG | INRIA | UGA

Citation

Charles Bouveyron, Stephane Girard, Cordelia Schmid. Classification of high dimensional data: High Dimensional Discriminant Analysis. Subspace, Latent Structure and Feature Selection techniques: Statistical and Optimisation perspectives Workshop, Feb 2005, Bohinj, Slovenia. 2005. 〈inria-00548517〉

Partager

Métriques

Consultations de la notice

363

Téléchargements de fichiers

297