Tight convex relaxations for sparse matrix factorization - Bio-informatique (CBIO) Accéder directement au contenu
Pré-Publication, Document De Travail Année : 2014

Tight convex relaxations for sparse matrix factorization

Résumé

Based on a new atomic norm, we propose a new convex formulation for sparse matrix factorization problems in which the number of nonzero elements of the factors is assumed fixed and known. The formulation counts sparse PCA with multiple factors, subspace clustering and low-rank sparse bilinear regression as potential applications. We compute slow rates and an upper bound on the statistical dimension of the suggested norm for rank 1 matrices, showing that its statistical dimension is an order of magnitude smaller than the usual $\ell_1$-norm, trace norm and their combinations. Even though our convex formulation is in theory hard and does not lead to provably polynomial time algorithmic schemes, we propose an active set algorithm leveraging the structure of the convex problem to solve it and show promising numerical results.
Fichier principal
Vignette du fichier
techreport_v2.pdf (702.49 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-01025959 , version 1 (18-07-2014)
hal-01025959 , version 2 (19-11-2014)
hal-01025959 , version 3 (04-12-2014)

Identifiants

Citer

Emile Richard, Guillaume Obozinski, Jean-Philippe Vert. Tight convex relaxations for sparse matrix factorization. 2014. ⟨hal-01025959v3⟩
830 Consultations
501 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More