Line search for generalized alternating projections

Mattias Fält, Pontus Giselsson

Research output: Chapter in Book/Report/Conference proceedingPaper in conference proceedingpeer-review

Abstract

This paper is about line search for the generalized alternating projections (GAP) method. This method is a generalization of the von Neumann alternating projections method, where instead of alternating projections, relaxed projections are alternated. The method can be interpreted as an averaged iteration of a nonexpansive mapping. Therefore, a recently proposed line search method for such algorithms is applicable to GAP. We evaluate this line search and show situations when the line search can be performed with little additional cost. We also present a variation of the basic line search for GAP - The projected line search. We prove its convergence and show that the line search condition is convex in the step length parameter. We show that almost all convex optimization problems can be solved using this approach and numerical results show superior performance with both the standard and the projected line search, sometimes by several orders of magnitude, compared to the nominal method.

Original languageEnglish
Title of host publication2017 American Control Conference, ACC 2017
PublisherIEEE - Institute of Electrical and Electronics Engineers Inc.
Pages4637-4642
Number of pages6
ISBN (Electronic)9781509059928
DOIs
Publication statusPublished - 2017 Jun 29
Event2017 American Control Conference, ACC 2017 - Seattle, United States
Duration: 2017 May 242017 May 26

Conference

Conference2017 American Control Conference, ACC 2017
Country/TerritoryUnited States
CitySeattle
Period2017/05/242017/05/26

Subject classification (UKÄ)

  • Computational Mathematics

Fingerprint

Dive into the research topics of 'Line search for generalized alternating projections'. Together they form a unique fingerprint.

Cite this