论文部分内容阅读
将遗传算法（ＧＡ）用于对ＢＰ网络权重的优化，不必对其进行参数调整，减少了训练的次数和计算量，且不会造成局部最优，多数影响遗传优化效果的规律得以确定、说明ＧＡ对神经网络的优化是可行的。
The genetic algorithm (GA) is used to optimize the weight of BP network. It is unnecessary to adjust the parameters and reduce the number of training and calculation, and will not cause local optimization. Most of the rules that affect the effect of genetic optimization are determined Optimization of neural network is feasible.