New Global Interchange Formula of Parameter Conjugate gradient method for solving optimization

Volume 27 , Issue 2 , December 2025 , Pages 12-22

Authors

Lajan Mohammed 1

1 University of Salahaddin, College of Science, Computer Science and INformation Technology Department

DOI logo 10.17656/sujpas.1014

Keywords

Abstract

In this study, we develop a new hybrid conjugate gradient method designed to generate a descent search direction that is independent of line search, thereby ensuring global convergence. We introduce an innovative interchange formula for the parameter, which consistently produces a descent direction for the objective function at each iteration. Through extensive numerical experiments, we demonstrate that the proposed algorithm achieves global convergence, performing comparably to other leading algorithms in the field. The results highlight the robustness and efficiency of our method, making it a valuable tool for unconstrained optimization problems.

Statistics
  • Article view442
  • Downloads28
  • First online25 December 2025
  • Published at25 December 2025

  • RIS
  • BibTeX
  • EndNote
  • Mendeley
  • APA (7th edition)
  • MLA (9th edition)
  • Chicago
  • Harvard
  • IEEE
  • Vancouver