Clustering Model Based on Dynamic Fuzzy Kohonen Network and Its Application
Yanling Yang · Journal of Nanchang University · 2010
The paper proposes a clustering model based on dynamic fuzzy Kohonen neural network.In view of the traditional Kohonen neural network(KNN) requiring a pre-determined number of clustering,it brings forward the tree dynamic self-organizing map neural network(TGSOM) to determine the cluster number,and optimizes network with using the advantage that the fuzzy Kohonen neural network(FKNN) clustering result does not depend the input sequence,to get better clustering results.The algorithm is applied to mining user interest model and proved its effectiveness by data test analysis.