Sample dispersion is better than sample discrepancy for classification - Archive ouverte HAL Accéder directement au contenu
Rapport Année : 2010

Sample dispersion is better than sample discrepancy for classification

Résumé

We want to generate learning data within the context of active learning. First, we recall theoretical results proposing discrepancy as a criterion for generating sample in regression. We show surprisingly that theoretical results about low discrepancy sequences in regression problems are not adequate for classification problems. Secondly we propose dispersion as a criterion for generating data. Then, we present numerical experiments which have a good degree of adequacy with theory.
Fichier principal
Vignette du fichier
RR-10-17.pdf (812.59 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-00679061 , version 1 (14-03-2012)

Identifiants

  • HAL Id : hal-00679061 , version 1

Citer

Benoît Gandar, Gaëlle Loosli, Guillaume Deffuant. Sample dispersion is better than sample discrepancy for classification. 2010. ⟨hal-00679061⟩
138 Consultations
227 Téléchargements

Partager

Gmail Facebook X LinkedIn More