Skip to Main content Skip to Navigation
Reports

Global linear convergence of an augmented Lagrangian algorithm for solving convex quadratic optimization problems

Abstract : We consider an augmented Lagrangian algorithm for minimizing a convex quadratic function subject to linear inequality constraints. Linear optimization is an important particular instance of this problem. We show that, provided the augmentation parameter is large enough, the constraint value converges globally linearly to zero. This property is proven by establishing first a global radial Lipschitz property of the reciprocal of the dual function subgradient. It is also a consequence of the proximal interpretation of the method. No strict complementarity assumption is needed. The result is illustrated by numerical experiments and algorithmic implications are discussed.
Document type :
Reports
Complete list of metadata

Cited literature [1 references]  Display  Hide  Download

https://hal.inria.fr/inria-00071556
Contributor : Rapport de Recherche Inria Connect in order to contact the contributor
Submitted on : Tuesday, May 23, 2006 - 5:56:53 PM
Last modification on : Friday, May 25, 2018 - 12:02:03 PM
Long-term archiving on: : Sunday, April 4, 2010 - 10:25:03 PM

Identifiers

  • HAL Id : inria-00071556, version 1

Collections

Citation

Frédéric Delbos, Jean Charles Gilbert. Global linear convergence of an augmented Lagrangian algorithm for solving convex quadratic optimization problems. [Research Report] RR-5028, INRIA. 2003. ⟨inria-00071556⟩

Share

Metrics

Record views

314

Files downloads

402