摘要
本文提出了一种新的模糊Kohonen聚类网络，它将模糊c-均值作为模糊控制机制结合到Kohonen网络学习和更新策略中，有效地解决了经典Kohonen聚类网络中学习参数启发式调整的缺点，仿真结果表明网络的收敛性和分类结果都得到了很大改善．
A new fuzzy Kohonen clustering network is proposed which integrates Fuzzy c-Means(FCM)model as the fuzzy control mechanism into the learning rate and updating strategies of Kohonen network.It effec tively overcomes the shortcomings which occur during clustering network parameters are adjusted heuristically The numerical results show improved convergence as well as reduced clustering errors.
出处
《电子学报》
EI
CAS
CSCD
北大核心
1998年第3期117-119,共3页
Acta Electronica Sinica
基金
国防预研基金
