Global convergence of the nonmonotone MBFGS method for nonconvex unconstrained minimization

被引：16
|
作者：
Zhou, Weijun [1 ]
Zhang, Li [1 ]
机构：
[1] Changsha Univ Sci & Technol, Coll Math & Computat Sci, Changsha 410076, Hunan, Peoples R China
来源：
JOURNAL OF COMPUTATIONAL AND APPLIED MATHEMATICS | 2009年 / 223卷 / 01期
基金：
美国国家科学基金会;
关键词：
Nonmonotone line search; MBFGS method; Nonconvex minimization; LINE SEARCH TECHNIQUE; NEWTON METHODS; BFGS METHOD;
D O I：
10.1016/j.cam.2007.12.011
中图分类号：
O29 [应用数学];
学科分类号：
070104 ;
摘要：
In this paper, we propose it new nonmonotone Armijo type line search and prove that the MBFGS method proposed by Li and Fukushima with this new fine search converges globally for nonconvex minimization. Some numerical experiments show that this nonmonotone MBFGS method is efficient for the given test problems. (C) 2007 Elsevier B.V. All rights reserved.
引用
收藏
页码：40 / 47
页数：8
相关论文
共 50 条
12345