An accelerated conjugate gradient method with adaptive two-parameter with applications in image restoration

被引：1

作者：

Zhu, Zhibin ^{[1
,2
]}

Zhu, Xiaowen ^{[1
,2
]}

Tan, Zhen ^{[1
,2
]}

机构：

[1] Guilin Univ Elect Technol, Guangxi Coll & Univ Key Lab Data Anal & Computat, Sch Math & Comp Sci, Guilin 541002, Peoples R China

[2] GUET, Ctr Appl Math Guangxi, Guilin 541002, Peoples R China

来源：

COMPUTATIONAL & APPLIED MATHEMATICS | 2024年 / 43卷 / 03期

基金：

中国国家自然科学基金;

关键词：

Sufficient descent condition; Powell restart strategy; Unconstrained optimization; Image restoration; Global convergence; CONVERGENCE CONDITIONS; ALGORITHMS; MINIMIZATION; DESCENT;

D O I：

10.1007/s40314-023-02521-5

中图分类号：

O29 [应用数学];

学科分类号：

070104 ;

摘要：

This paper proposes an adaptive two-parameter accelerated conjugate gradient method, which satisfies the sufficient descent condition in the search direction. The Powell restart strategy is designed in the algorithm to improve its numerical performance. Our proposed method does not add extra computational effort compared to other methods. Furthermore, under general assumptions, we demonstrate the global convergence of our proposed method under the Wolfe line search. Finally, we compare with other methods on the unconstrained optimization and image restoration problems. Numerical experiments are presented to show that our proposed method is feasible.

引用

页数：20