参数可调的自扩展聚类算法及其应用  

Parameter-adjustable self-expanded clustering algorithm and its application

作　　者：张洪涛[1] 张坤[2] 马培军[2] 

机构地区：[1]哈尔滨工业大学航天学院,哈尔滨150001 [2]哈尔滨工业大学计算机科学与技术学院,哈尔滨150001

出　　处：《哈尔滨工业大学学报》2007年第11期1695-1698,共4页Journal of Harbin Institute of Technology

基　　金：国家自然科学基金资助项目(60273083)

摘　　要：针对大规模样本聚类的时间复杂度过高和聚类结果对经验参数设置的依赖性过强的问题,提出一种基于密度网格可变参数的自扩展聚类算法.算法将数据空间分割成相同大小的网格单元,再将样本归一化后映射到相应的网格单元中,然后从指定密度较大的网格单元向周围扩展,直到其平均密度达到指定的下限或可扩展的聚类边界为止.聚类过程中,通过下限密度和均值密度限制聚类间的过度扩展,如果有效样本的比率低于阈值,则自适应调整扩展密度并重新聚类.仿真试验表明,本算法可以以较小的时间代价获得较高的聚类精度和有效样本率.Aiming at the problem that too much time of the large scale samples clustering and the result of the clustering excessively relies on the experiential parameters, an efficient self-expanded clustering algorithm with parameter-adjustable based on density units（PASCDU） is proposed in this paper. The whole data space is di- vided into several equal density units, before each data point is mapped into the relevant density unit according to the data point charter. Then cluster extends around from large density unit, until average density under the low-limit or cluster extends to the cluster＇ s edge. During the clustering process, we can prevent clusters from excessively extending by setting low-limit and average-limit. If the ratio of effectual samples is under the designate value, then automatically adjust this value, and then cluster again. The experiment indicates that this algorithm can expend short time to improve the cluster precision and the ratio of effectual samples.

关 键 词：自扩展聚类 密度网格 模糊神经网络 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]