一种改进的层次聚类算法

An Improved Algorithm of Hierarchical Clustering

为了更好地实现聚类,在分析层次聚类(agglomerative)算法和神经网络的ART2算法的基础上,提出了一种改进的层次聚类算法.改进算法将首先采用一种基于ART2的改进神经网络聚类算法得到一个初始的聚类结果,然后在此基础上利用agglomerative算法实现分层聚类.实验结果表明,改进算法较原先传统的聚类算法,不但算法执行速度快、效率高,而且聚类效果也比较好.

In order to achieve clustering well,a modified hierarchical clustering algorithm is proposed based on the strengths and weaknesses of hierarchical clustering（agglomerative） algorithm and neural network ART2 algorithm.Improved algorithm will first use an improved ART2 clustering algorithm to form initial clustering results,and then achieve hierarchical clustering result by agglomerative clustering algorithm based on the results of the previous.It is proved that the proposed algorithm is not only faster than the traditional clustering algorithm,but also the clustering result is better.