论文部分内容阅读
基于粒子群优化的算法具有全局随机搜索最优解的特点。本文尝试把PSO算法和神经网络权值训练的常用算法BP算法结合起来进行数据的训练,实现对一组数据的训练,并对结果与BP算法的训练结果进行了对比,得到了较好的效果。
Particle swarm optimization algorithm has the characteristics of global optimal search for random search. This paper attempts to PSO algorithm and neural network weights training common algorithm BP algorithm combined data training, to achieve a set of data training, and the results with the BP algorithm training results were compared, and achieved good results .