简单下降梯度法

Simple Decent Gradient Methods

构造了一解无约束优化问题算法 ,与其它算法比较 ,最主要不同在于此类算法每次迭代只需简单下降 ,不需要充分下降 在一般性假设下 ,算法是收敛的

This paper presents an algorithm for unconstrained optimizat io n.Compared with other similar algorithm,it has a characteristic needless suffi cient decent.Under common assumption,it is convergent.The numerical experimen t shows it is valid.