Article Open Access 10.26855/jamc.2018.09.002
New Hybrid Conjugate Gradient Method as A Convex Combination of HS and FR Methods
Snezana S. Djordjevic
1Faculty of Technology, University of Nis, 16000 Leskovac, Serbia
*Corresponding author: Snezana S. Djordjevic
Email: snezanadjordjevic1971@gmail.com
Published: September 27,2018
Abstract
In this paper we present a new hybrid conjugate gradient
algorithm for unconstrained optimization. This method is a convex combination
of Hestenes-Stiefel conjugate gradient method and Fletcher-Reeves conjugate
gradient method. The parameter is chosen in such a
way that the search direction satisfies the condition of the Newton direction.
The strong Wolfe line search conditions are used. The global convergence of new
method is proved.
Numerical comparisons show that the present hybrid conjugate
gradient algorithm is the efficient one.
How to cite this paper
New Hybrid Conjugate Gradient Method as A Convex Combination of HS and FR Methods
How to cite this paper: Snezana, S.D. (2018) New Hybrid Conjugate Gradient Method as A Convex Combination of HS and FR Methods. Journal of Applied Mathematics and Computation, 2(9), 366-378.
DOI: 10.26855/jamc.2018.09.002