A MODIFIED CONJUGATE-DESCENT METHOD AND ITS GLOBAL CONVERGENCE

被引：0
|
作者：
Li, Min [2 ]
Li, Dong-Hui [1 ]
机构：
[1] S China Normal Univ, Sch Math Sci, Guangzhou 510631, Guangdong, Peoples R China
[2] Huaihua Univ, Dept Math & Appl Math, Huaihua 418008, Peoples R China
来源：
PACIFIC JOURNAL OF OPTIMIZATION | 2012年 / 8卷 / 02期
关键词：
unconstrained optimization; Conjugate-Descent method; sufficient descent direction; global convergence; GRADIENT METHOD; LINE SEARCH; GUARANTEED DESCENT; ASCENT METHODS; OPTIMIZATION; ALGORITHM;
D O I：
暂无
中图分类号：
C93 [管理学]; O22 [运筹学];
学科分类号：
070105 ; 12 ; 1201 ; 1202 ; 120202 ;
摘要：
In this paper, we make a modification to the Conjugate-Descent (CD) method and propose a descent CD method. The method can generate sufficient descent directions for the objective function. Under mild conditions we show that the method is globally convergent if the Armijo line search or the standard Wolfe line search is used. The numerical results show that the proposed descent CD method is efficient.
引用
收藏
页码：247 / 259
页数：13
相关论文
共 50 条
12345