“鞍面法”实为一个鞍点法

“Saddle Surface Algorithm”in Practice is a Saddle Point Algorithm

本文根据[3][4]中求解线性规划鞍点法的思想,将所谓“鞍面法”的迭代公式进行简化,并指出修正后的算法实际上是一个总体线性收敛的鞍点法。

In this paper,according to the idea about saddle point algorithm for linear programming in [3] [4], we simplied the iterative formula of the so called'saddle surface algorithm'and pointed out,that the modified algorithm in practice is a globally linearly convergent saddle point algorithm.