The degrees of freedom of the Lasso in underdetermined linear regression models

Abstract : In this paper, we investigate the degrees of freedom (df) of penalized l1 minimization (also known as the Lasso) for an un-derdetermined linear regression model. We show that under a suitable condition on the design matrix, the number of nonzero coefficients of the Lasso solution is an unbiased estimate for the degrees of freedom. An effective estimator of the number of degrees of freedom may have several applications including an objectively guided choice of the regularization parameter in the Lasso through the SURE or GCV frameworks.
Type de document :
Communication dans un congrès
SPARS 2011, Jul 2011, Edinburgh, United Kingdom. pp.56, 2011
Liste complète des métadonnées

https://hal.archives-ouvertes.fr/hal-00625219
Contributeur : Gabriel Peyré <>
Soumis le : mercredi 21 septembre 2011 - 10:14:22
Dernière modification le : mercredi 28 septembre 2016 - 16:14:49
Document(s) archivé(s) le : mardi 13 novembre 2012 - 14:06:26

Fichier

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

Identifiants

  • HAL Id : hal-00625219, version 1

Collections

Citation

Maher Kachour, Jalal M. Fadili, Christophe Chesneau, Charles Dossal, Gabriel Peyré. The degrees of freedom of the Lasso in underdetermined linear regression models. SPARS 2011, Jul 2011, Edinburgh, United Kingdom. pp.56, 2011. <hal-00625219>

Partager

Métriques

Consultations de
la notice

303

Téléchargements du document

117