Eigenrange: A Robust Spectral Method for Dimensionality Reduction

Abstract : This paper addresses the problem of dimension reduction of noisy data, more precisely the challenge of determining the dimension of the subspace where the signal lives in. Based on results from random matrix theory, a novel estimator of the signal dimension is developed. Consistency of the estimator is proved in the modern asymptotic regime, where the number of parameters grows proportionally with the sample size. Experimental results show that the novel estimator is robust to noise and, moreover, it gives highly accurate results in settings where standard methods fail. The application of the new dimension estimator on several biomedical data sets in the context of classification illustrates the improvements achieved by the new method compared to the state of the art.
Type de document :
Pré-publication, Document de travail
2016
Liste complète des métadonnées

Littérature citée [30 références]  Voir  Masquer  Télécharger

https://hal.archives-ouvertes.fr/hal-01400952
Contributeur : Tabea Rebafka <>
Soumis le : mardi 22 novembre 2016 - 16:33:38
Dernière modification le : lundi 29 mai 2017 - 15:33:35
Document(s) archivé(s) le : lundi 20 mars 2017 - 20:30:08

Identifiants

  • HAL Id : hal-01400952, version 1

Collections

UPMC | ICAN | PMA | USPC

Citation

Malika Kharouf, Tabea Rebafka, Nataliya Sokolovska. Eigenrange: A Robust Spectral Method for Dimensionality Reduction. 2016. 〈hal-01400952〉

Partager

Métriques

Consultations de
la notice

183

Téléchargements du document

224