A New Dai-Liao Type of Conjugate Gradient Algorithm for Unconstrained Optimization Problems

Release time:2019-03-01

Affiliation of Author(s):中南大学

Journal:Journal of Pacific Optimization

Place of Publication:Japan

Key Words:unconstrained optimization; three-term conjugate gradient method; global convergence; Armijo-type

Abstract:In this paper, a new Dai-Liao type of three-term conjugate gradient algorithm is developed for solving nonconvex unconstrained optimization problems. The search direction consists of three terms, which aim to gather more useful information of the current iterate point such that the direction has better convergence performance for the algorithm. Different from the existing methods, global convergence is established without assumption of uniform convexity under a modified Armijo-type line search. Numerical experiments are employed to show efficiency of the algorithm in solving large-sca

Co-author:Jing Lv

First Author:Songhai Deng

Indexed by:Unit Twenty Basic Research

Correspondence Author:Zhong Wan

Document Type:J

ISSN No.:1349-8169

Translation or Not:no

Date of Publication:2018-11-09