摘要
对模糊ART神经网络模型中的类别选择方法进行改进,并在权值向量的修改规则中引入隶属度,得到一种改进的FuzzyART学习算法。IRIS数据分类结果证明了新方法的可行性。
In this paper,we improve the cluster-choose way of Fuzzy ART neural network,and moreover introduce the membership degree into the updating rule of connection weight vector,and then gain an improved Fuzzy ART learning algorithm. IRIS data classification results prove the validity of the new algorithm.
出处
《计算机工程与应用》
CSCD
北大核心
2008年第28期49-50,85,共3页
Computer Engineering and Applications
关键词
模糊自适应共振理论
隶属度
模式分类
Fuzzy Adaptive Resonance Theory
membership degree
pattern classification
