Multiple Indefinite Kernel Learning with Mixed Norm Regularization - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2009

Multiple Indefinite Kernel Learning with Mixed Norm Regularization

Résumé

We address the problem of learning classifiers using several kernel functions. On the contrary to many contributions in the field of learning from different sources of information using kernels, we here do not assume that the kernels used are positive definite. The learning problem that we are interested in involves a misclassification loss term and a regularization term that is expressed by means of a mixed norm. The use of a mixed norm allows us to enforce some sparsity structure, a particular case of which is, for instance, the Group Lasso. We solve the convex problem by employing proximal minimization algorithms, which can be viewed as refined versions of gradient descent procedures capable of natu- rally dealing with nondifferentiability. A numerical simulation on a UCI dataset shows the modularity of our approach.
Fichier principal
Vignette du fichier
mixednormik.pdf (210 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-00424033 , version 1 (14-10-2009)

Identifiants

  • HAL Id : hal-00424033 , version 1

Citer

Matthieu Kowalski, Marie Szafranski, Liva Ralaivola. Multiple Indefinite Kernel Learning with Mixed Norm Regularization. International Conference on Machine Learning (ICML 2009), Jun 2009, Montreal, Canada. pp.520. ⟨hal-00424033⟩
947 Consultations
234 Téléchargements

Partager

Gmail Facebook X LinkedIn More