The convergence properties of new hybrid conjugate gradient method

3rd Indonesian Operations Research Association - International Conference on Operations Research 2018

计算机科学

Salih, Y.^1 ; Hamoda, M.A.^2 ; Sukono^3 ; Mamat, M.^1

Department of Computer Science and Mathematics, Faculty of Informatics and Computing, Universiti Sultan Zainal Abidin, Terengganu

22200, Malaysia^1

Department of Mathematics, Faculty of Science, AlAsmarya Islamic University, Libya^2

Universitas Padjadjaran, Bandung, Indonesia^3

关键词: Convergence properties;    Exact line searches;    Global convergence properties;    Hybrid method;    Large scale nonlinear optimization;    Numerical experiments;    Search direction;    Test problem;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/567/1/012031/pdf
DOI  :  10.1088/1757-899X/567/1/012031

学科分类：计算机科学（综合）

来源: IOP

【 摘 要 】

In this paper, a new Hybrid Conjugate Gradient Methods is presented, which produced sufficient descent search direction at every iteration and global Convergence Properties for solving large-scale nonlinear optimization problem under exact line search. The numerical experiments show that a hybrid method has the best efficiency for the test problems.