中文摘要：

针对传统的划分聚类算法不能够发现任意形状的簇的缺点,引入一种能够有效反映样本间相似度的距离度量——基于路径的距离度量,并设计了一种能够反映类内样本相似度大、类间样本相似度小的目标准则函数.实验表明,本文提出的基于路径划分的聚类算法与传统的k均值算法相比具有更好的聚类效果.

英文摘要：

Traditional partition clustering algorithm can＇t discover clusters of arbitrary shapes.For this problem,a new path-based similarity measure is proposed,which can reflect the similarity between samples effectively.A new objective criterion function is designed which can show that one sample is more similar to another from the same cluster than that from a different cluster.Experimental results show that the proposed method can get better clustering results than k-means algorithm.