Proximal Algorithm Meets a Conjugate descent
Résumé
An extansion of the non linear conjugate gradient algorithm is proposed for some non smooth problems. We extend some results of descent algorithm in the smooth case for convex non smooth functions. We then construct a conjugate descente algorithm based on the proximity operator to obtain a descent direction. Analysis of convergence of this algorithme is provided.
Origine | Fichiers produits par l'(les) auteur(s) |
---|