Variable metric forward-backward algorithm for minimizing the sum of a differentiable function and a convex function

Abstract : We consider the minimization of a function $G$ defined on $R^N$, which is the sum of a (non necessarily convex) differentiable function and a (non necessarily differentiable) convex function. Moreover, we assume that $G$ satisfies the Kurdyka-Lojasiewicz property. Such a problem can be solved with the Forward-Backward algorithm. However, the latter algorithm may suffer from slow convergence. We propose an acceleration strategy based on the use of variable metrics and of the Majorize-Minimize principle. We give conditions under which the sequence generated by the resulting Variable Metric Forward-Backward algorithm converges to a critical point of $G$. Numerical results illustrate the performance of the proposed algorithm in an image reconstruction application.
Type de document :
Article dans une revue
Journal of Optimization Theory and Applications, Springer Verlag, 2014, 162 (1), pp.107-132. <10.1007/s10957-013-0465-7>
Liste complète des métadonnées


https://hal.archives-ouvertes.fr/hal-00789970
Contributeur : Emilie Chouzenoux <>
Soumis le : mardi 19 février 2013 - 10:38:01
Dernière modification le : jeudi 10 juillet 2014 - 11:01:43
Document(s) archivé(s) le : lundi 20 mai 2013 - 04:00:19

Fichier

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

Identifiants

Citation

Emilie Chouzenoux, Jean-Christophe Pesquet, Audrey Repetti. Variable metric forward-backward algorithm for minimizing the sum of a differentiable function and a convex function. Journal of Optimization Theory and Applications, Springer Verlag, 2014, 162 (1), pp.107-132. <10.1007/s10957-013-0465-7>. <hal-00789970>

Partager

Métriques

Consultations de
la notice

432

Téléchargements du document

391