混合模糊谱系聚类算法

Fuzzy hybrid hierarchical clustering algorithm

针对模糊C-均值聚类算法对聚类数预先不可知和谱系聚类所具有的缺陷,提出了混合模糊谱系聚类算法,该算法结合模糊聚类和谱系聚类,自动确定聚类数目,并可以有效的对数据进行聚类.实验表明,该算法具有良好的有效性和可行性.

Fuzzy C-means algorithm requires the user to predefine the number of clusters （c） ; however, it is not always possible to know c a priori, and classic hierarchical method has several deficiencies. Aiming at above problem,this paper propose hybrid fuzzy hierarchical clustering algorithm, which combines fuzzy clustering and hierarchical clustering to find the number of clusters and cluster. Test shows the proposed algorithm has superior effectiveness and feasibility.