Restarting the accelerated coordinate descent method with a rough strong convexity estimate - Télécom Paris Accéder directement au contenu
Article Dans Une Revue Computational Optimization and Applications Année : 2018

Restarting the accelerated coordinate descent method with a rough strong convexity estimate

Résumé

We propose new restarting strategies for the accelerated coordinate descent method. Our main contribution is to show that for a well chosen sequence of restarting times, the restarted method has a nearly geometric rate of convergence. A major feature of the method is that it can take profit of the local quadratic error bound of the objective function without knowing the actual value of the error bound. We also show that under the more restrictive assumption that the objective function is strongly convex, any fixed restart period leads to a geometric rate of convergence. Finally, we illustrate the properties of the algorithm on a regularized logistic regression problem and on a Lasso problem.
Fichier non déposé

Dates et versions

hal-02287842 , version 1 (13-09-2019)

Identifiants

  • HAL Id : hal-02287842 , version 1

Citer

Olivier Fercoq, Zheng Qu. Restarting the accelerated coordinate descent method with a rough strong convexity estimate. Computational Optimization and Applications, 2018. ⟨hal-02287842⟩
39 Consultations
0 Téléchargements

Partager

Gmail Facebook X LinkedIn More