摘要
文章首先概述了指标选择方法的现状,并指出现有方法在全面性和代表性上的问题。接着提出了一种用聚类分析、非参数检验等构成的指标动态筛选方法,有效的解决了全面性和代表性的问题。最后通过实例说明了该方法的应用。
First we illuminate the basal principles about index's choose in evaluation and point out that all of the methods now have some questions about representation and omnipotence. Then a nwe method, composed by clustering analysis and non-parameter test and so on, is put forward, which settles the defects between representation and omnipotence. Using it, a simple example is shown to us.
出处
《数理统计与管理》
CSSCI
北大核心
2004年第1期72-76,共5页
Journal of Applied Statistics and Management
