摘要
提出一种新的基于免疫的RBF网络在线算法.首先是融入增加结点策略和减少结点策略,把传统的免疫RBF网络算法改进成在线学习算法.其次是改进了权值学习算法,径向基函数相当于这一类的概率密度,隐层到输出层权值相当于这一类的值.用这种方法权值不需要训练.试验结果表明,该方法效果理想、速度快,识别率高.
Put forward a new network of RBF of based on the immunity,which is a on-line algorithm.The first is adding strategy and remarkable degree based on the output of radial basis function which is a pruning strategy.The second is improval of the weight value study algorithm.radial basis function is equal to probability of this type,weight is equal to value of this type.Weight doesn t need to train.Experiment results show that this method is ideal,speed is quick,and detection rate is high.
出处
《福州大学学报(自然科学版)》
CAS
CSCD
北大核心
2008年第S1期27-30,共4页
Journal of Fuzhou University(Natural Science Edition)
基金
国家自然科学基金资助项目(60675058)
福建省自然科学基金资助项目(A0610013
A0710008)