A semismooth newton method with multidimensional filter globalization for l1-optimization

Andre Milzarek, Michael Ulbrich

Publikation: Beitrag in FachzeitschriftArtikelBegutachtung

48 Zitate (Scopus)

Abstract

Due to their property of enhancing the sparsity of solutions, l 1-regularized optimization problems have developed into a highly dynamic research area with a wide range of applications. We present a class of methods for l1-regularized optimization problems that are based on a combination of semismooth Newton steps, a filter globalization, and shrinkage/thresholding steps. A multidimensional filter framework is used to control the acceptance and to evaluate the quality of the semismooth Newton steps. If the current Newton iterate is rejected a shrinkage/thresholdingbased step with quasi-Armijo stepsize rule is used instead. Global convergence and transition to local q-superlinear convergence for both convex and nonconvex objective functions are established. We present numerical results and comparisons with several state-of-the-art methods that show the efficiency and competitiveness of the proposed method.

OriginalspracheEnglisch
Seiten (von - bis)298-333
Seitenumfang36
FachzeitschriftSIAM Journal on Optimization
Jahrgang24
Ausgabenummer1
DOIs
PublikationsstatusVeröffentlicht - 2014

Fingerprint

Untersuchen Sie die Forschungsthemen von „A semismooth newton method with multidimensional filter globalization for l1-optimization“. Zusammen bilden sie einen einzigartigen Fingerprint.

Dieses zitieren