摘要
针对RBF网络易于陷入局部最小值和网络训练时间过长的问题,引入免疫多克隆遗传算子对RBF网络的权值进行优化。仿真结果表明,采用优化后的网络模型具有更高的分类效率和准确率。
P.BF networks are easy to get into a local minimum and time, plyctonal genetic operators to introduce immune RBF network show that optimize the use of network model with higher efficiency network training problem of excessive weights optimizing. Simulations results and accuracy of classification.
出处
《柳州职业技术学院学报》
2013年第3期32-36,共5页
Journal of Liuzhou Vocational & Technical College
关键词
免疫算法
多克隆算法
RBF神经网络
DNA序列分类
immune algorithm,poly-clonal algorithm,RBF neural network,DNA sequence classification