基于模糊聚类的数据挖掘方法与应用

Research and implementing of data mining method based on fuzzy clustering

系统聚类法中常用的是最小距离法、最大距离法、重心距离法和类平均距离法等,这些方法都是定义一种类与类之间的距离来进行聚类的,但在有些情况下其聚类结果不唯一。利用模糊关系矩阵,给出一种新的基于模糊聚类的方法,并将这些技术应用到处理地质图型数据的数据挖掘工作中,得到可行性验证。

System clustering had methods of least distance, largest distance, baryeenter distance, class average distance, and so on.These methods were clustered by distance between class and class, but its result was not unique under some cases. It was used fuzzy matrix and presented a kind of fuzzy clustering method. These technologies were applied into data mining of geologic graphics data dealing and the feasibility of data mining had been tested by experiments successfully.