With the clustering feature analyzed, restrained function and pattern similarity are introduced. Then the algorithm of improved FCM is presented.
通过对模糊c均值算法聚类特性的分析,引入了约束函数及模式相似度的概念,提出了改进的FCM算法。
The algorithm divides tag co -occurrence network based on tag node's centrality and similarity, and automatically generates a cluster feature tag after clustering to represent that cluster.
算法基于标签节点的核心度和相似性对标签共现网络进行分割,并在聚类后自动生成该类的特征标签来代表该类簇。
In this paper, the basic principle of the clustering algorithm based on self-organizing feature map network is discussed, and pointed out its defects.
本文讨论了基于自组织特征映射网络聚类算法的基本原理,并指出了算法的缺陷。
应用推荐