Numerical Algorithms

, Volume 59, Issue 1, pp 79–93

Global convergence of a modified Hestenes-Stiefel nonlinear conjugate gradient method with Armijo line search

Authors

    • College of Mathematics and Computational ScienceChangsha University of Science and Technology
  • Fenghua Wen
    • School of Econometrics and ManagementChangsha University of Science and Technology
Original Paper

DOI: 10.1007/s11075-011-9477-2

Cite this article as:
Dai, Z. & Wen, F. Numer Algor (2012) 59: 79. doi:10.1007/s11075-011-9477-2

Abstract

In this article, based on the modified secant equation, we propose a modified Hestenes-Stiefel (HS) conjugate gradient method which has similar form as the CG-DESCENT method proposed by Hager and Zhang (SIAM J Optim 16:170–192, 2005). The presented method can generate sufficient descent directions without any line search. Under some mild conditions, we show that it is globally convergent with Armijo line search. Moreover, the R-linear convergence rate of the modified HS method is established. Preliminary numerical results show that the proposed method is promising, and competitive with the well-known CG-DESCENT method.

Keywords

Unconstrained optimizationConjugate gradient methodSufficient descent propertyR-linear convergenceGlobal convergence

Mathematics Subject Classifications (2010)

90C3065K05

Copyright information

© Springer Science+Business Media, LLC. 2011