Globally Convergent Three-Term Conjugate Gradient Methods that Use Secant Conditions and Generate Descent Search Directions for Unconstrained Optimization

Kaori Sugiki, Yasushi Narushima, Hiroshi Yabe

Research output: Contribution to journalArticlepeer-review

69 Citations (Scopus)

Abstract

In this paper, we propose a three-term conjugate gradient method based on secant conditions for unconstrained optimization problems. Specifically, we apply the idea of Dai and Liao (in Appl. Math. Optim. 43: 87-101, 2001) to the three-term conjugate gradient method proposed by Narushima et al. (in SIAM J. Optim. 21: 212-230, 2011). Moreover, we derive a special-purpose three-term conjugate gradient method for a problem, whose objective function has a special structure, and apply it to nonlinear least squares problems. We prove the global convergence properties of the proposed methods. Finally, some numerical results are given to show the performance of our methods.

Original languageEnglish
Pages (from-to)733-757
Number of pages25
JournalJournal of Optimization Theory and Applications
Volume153
Issue number3
DOIs
Publication statusPublished - 2012 Jun
Externally publishedYes

Keywords

  • Descent search direction
  • Global convergence
  • Secant condition
  • Three-term conjugate gradient method
  • Unconstrained optimization

ASJC Scopus subject areas

  • Management Science and Operations Research
  • Control and Optimization
  • Applied Mathematics

Fingerprint

Dive into the research topics of 'Globally Convergent Three-Term Conjugate Gradient Methods that Use Secant Conditions and Generate Descent Search Directions for Unconstrained Optimization'. Together they form a unique fingerprint.

Cite this