A convergent modified HS-DY hybrid conjugate gradient method for unconstrained optimization problems. Issue 1 (2nd January 2019)

• Record Type: Journal Article
• Authors: Mtagulwa, Peter; Kaelo, P.
• Is Part Of: Journal of information & optimization sciences. Volume 40:Issue 1(2019)
• Page Start: 97
• Page End: 113
• Publication Date: 2019-01-02
• Subjects: (2010) 90C30 -- 65K05 -- 90C06; Unconstrained optimization -- Global convergence -- Conjugate gradient method -- Sufficient descent -- Wolfe-Powell conditions; Electronic data processing -- Periodicals; Information science -- Periodicals; Mathematical optimization -- Periodicals; 519.6
• DOI: 10.1080/02522667.2018.1424087
• Languages: English
• ISSNs: 0252-2667

Abstract

Abstract: Conjugate gradient algorithm (method) is a very simple and powerful technique for solving large scale unconstrained optimization problems. In this paper a new modified hybrid conjugate gradient method, based on the work of Liu and Du [17] and Dong, Jiao and Chen [8], is proposed. We show that the proposed algorithm satisfies the descent condition and its global convergence is also established under the weak Wolfe-Powell line search conditions. The proposed algorithm is tested on a number of benchmark problems and the numerical results show that the proposed algorithm is very competitive.