Highly effective fuzzy clustering algorithm based on improved network
一种基于改进神经网络的高效模糊聚类算法

In order to solve the problems in fuzzy clustering by using Self-Organizing Feature Map (SOFM) network, this paper introduced an improved structural self-organizing feature map network and adopted self-adapting initial condition. It can handle the clustering problem of high dimensional data and the clusters with arbitrary shapes. Compared with other algorithms with the same clustering effect, it has lower clustering time complexity. Experiments indicate this algorithm has better clustering effect compared to single SOFM network and other kin algorithms.