|
| 摘要: |
| 本文研究了无约束最优化的求解问题.利用新的对角拟牛顿校正和非单调技术,获得了一种非单调广义对角拟牛顿算法.新算法具有低存储、低计算量的特点,非常适合大规模问题的求解,推广了文献[8]的结果. |
| 关键词: 弱拟牛顿方程 对角校正 非单调技术 全局收敛性 数值实验 |
| DOI： |
| 分类号:O224 |
| 基金项目:江苏省高校自然科学研究项目(13KJB110007);江苏理工学院基础及应用基础研究项目(KYY13012);江苏理工学院博士启动基金项目(KYY13005) |
|
| NONMONOTONE GENERALIZED DIAGONAL QUASI-NEWTON ALGORITHM |
|
ZHOU Qun-yan,HANG Dan
|
| Abstract: |
| In this paper, the method for solving unconstrained optimization is studied. By using the new diagonal quasi-Newton update and nonmonotone technique, a nonmonotone generalized diagonal quasi-Newton algorithm is proposed. The new method needs less memory capacitance and computational complexity. It is very effective and attractive for large scale unconstrained problems and generalizes the results of reference[8]. |
| Key words: weak quasi-Newton equation diagonal updating nonmonotone technique global convergence numerical experiment |
