Performance Analysis of New Spectral and Hybrid Conjugate Gradient Methods for Solving Unconstrained Optimization Problems

Maulana Malik, Mustafa Mamat, Siti Sabariah Abas, Ibrahim Mohammed Sulaiman, S. Sukono

Research output: Contribution to journalArticlepeer-review

Abstract

The spectral and hybrid conjugate gradient methods are part of the conjugate gradient methods. Conjugate gradient methods are among the iterative method for solving unconstrained optimization problems. In this paper, a new spectral and hybrid conjugate gradient methods are proposed. Based on some assumptions and strong Wolfe line search, the new spectral conjugate gradient method satisfies the global convergence properties. As well as the hybrid conjugate gradient method fulfill the global convergence properties under an exactline search. We also prove that the proposed methods fulfill the sufficient descent condition. Finally, based on some test problems, the numerical results of the proposed methods are very competitive and most efficient.

Original languageEnglish
JournalIAENG International Journal of Computer Science
Volume48
Issue number1
Publication statusPublished - 2021

Keywords

  • exact line search
  • global convergence properties
  • hybrid conjugate gradient method
  • spectral conjugate gradient method
  • Strong Wolfe line search
  • sufficient descent condition

Fingerprint Dive into the research topics of 'Performance Analysis of New Spectral and Hybrid Conjugate Gradient Methods for Solving Unconstrained Optimization Problems'. Together they form a unique fingerprint.

Cite this