数据挖掘聚类算法研究及实现

Research and Implementation of Clustering Algorithm of Data Mining

聚类算法是数据挖掘的核心技术。研究了一种多重系统聚类模型及其算法实现,将变量聚类和样本聚类结合起来,先将指标按一定的规则分成若干类别,然后对包含每一类别指标的样本数据分别进行聚类,使用这种多重聚类方法,分类性能有了较大提高。并且在此算法多重聚类的实现中,使用了两种方法赋值样本数据阵,使聚类的结果有了直观的比较。

Clustering algorithm is the core of data mining technology. In this paper, we propose a model of multilevel system clustering, which integrates Q-clustering with R-clustering, first take the sample and index dividing into the some category, then clustering the sample data that including each class sample and index. Using this kind of multilevel achieves a high performance. And then in this algorithm, using two methods to express sample data. Making the result of clustering has the comparison directly.