基于弱牛顿方程的一种无记忆方法

A New Memoryless Method Based on Weak Quasi-Newton Equation

给出了一种新的求解无约束优化问题的方法。为了求下一次迭代点时 ,它不需要进行矩阵计算。在不精确线搜索 (Armijo- Goldstein准则 )下 ,该方法对一般目标函数为全局收敛 ,且对一类特殊的凸函数 ,其定步长型算法亦是收敛的 ,并给出了计算的步骤和计算的数值例子。

This paper proves a new method for solving unconstrained optimization problems.Here for the next iterative point,it needn’t to do matrix calculus,under inexact line searches(Armijo Goldstein rule),this method is globally convergent when applied to a general objective function.Moreover,the fixed step algorithm is convergent for a special kind of convex objective fuctions.