基于粒子群优化的BP网络学习算法  

A BP NETWORKS LEARNING ALGORITHM BASED ON PSO

作　　者：王岁花[1] 冯乃勤[1] 李爱国[2] 

机构地区：[1]河南师范大学计算机科学系,新乡453002 [2]西安科技学院计算机科学系,西安710054

出　　处：《计算机应用与软件》2003年第8期74-76,共3页Computer Applications and Software

摘　　要：本文提出一种新颖的基于粒子群优化的BP网络学习算法 ,该算法是一种全局随机优化算法。用Iris分类问题 ,将所提出的算法与BP算法作了对比实验。实验结果表明 :所提出的算法性能优于BP算法 。A novel BP neural networks learning algorithm based on particle swarm optimizers (PSO)is proposed in this paper.A stochastic global optimization technique is used in proposed algorithm.Comparing experiments that classify Iris problem are made between proposed algorithm and BP algorithm.The experiment results show that proposed algorithm does not only superior to BP algorithm,bus also has fast congruence speed.

关 键 词：人工神经网络 BP网络 学习算法 粒子群优化 全局随机优化算法 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]