基于分类指标优化的聚类分析

Proceeding cluster analysis with the optimized index

针对当前应用聚类分析直接从现成指标入手的现状,指出了其局限性.应用多元统计分析,首先对初步建立的分类指标中关联性大的指标进行优化整合,得到一组更为科学合理的分类指标,再进行模糊聚类分析,聚类结果证明此思路是正确的.

The limitation of applying the cluster analysis with ready-made index is put forward according to the current condition .Applying multivariate analysis, the ready-made index with close connection is firstly optimized .Then fuzziness cluster analysis is proceeded with the more reasonable indexes which has been acquired from the optimized ones. The result shows that the direction of thought is right.