摘要
介绍了粒子群优化(PSO)算法的原理,研究了将PSO算法应用于神经网络训练的方法,给出了算法软件实现的基本流程,并对Iris分类问题做了仿真实验,通过与BP算法的比较,结果表明基于PSO的神经网络训练算法操作简单,易于实现,而且训练精度较高,有良好的收敛性.
In the paper, the fundamental principles of PSO are introduced firstly. And then, the processes of training neural network by applying PSO algorithm are presented. Simulating experiments of Iris classification problem are also made between proposed algorithm and BP algorithm. A comparison of the two algorithms indicates that the former is a simple one with good convergence.
出处
《河南师范大学学报(自然科学版)》
CAS
CSCD
北大核心
2007年第3期169-171,共3页
Journal of Henan Normal University(Natural Science Edition)
基金
河南省科技攻关项目(424300008)
关键词
粒子群优化
进化算法
神经网络
particle swarm optimization
evolutionary algorithm
neural network