摘要
目的讨论取值稠密信息系统的一种属性约简方法。方法通过对取值稠密信息系统进行预处理——离散化,把它们转化为经典的信息系统,然后以有向关联信息作为度量标准对转化后所得的经典信息系统进行属性约简。结果/结论该方法可对取值稠密信息系统有效地进行属性约简,使此类信息系统的规则表示简单直观。
Aim An attribute reduction method of dense domain information systems was discussed. Methods Dense domain information systems was inverted to classic information systems by preconditioning——discretization.And we gave a method of attribute reduction of classic information systems using oriented relevance information. Results/Conclusion The test shows that this approach not only succeeds in giving dense domain information systems attribute reduction,but also gets effective conclusions.
出处
《宝鸡文理学院学报(自然科学版)》
CAS
2009年第3期68-72,共5页
Journal of Baoji University of Arts and Sciences(Natural Science Edition)
基金
宝鸡文理学院重点项目(ZK08137)
关键词
取值稠密信息系统
有向关联信息
属性约简
dense domain information systems
oriented relevance information
attribute reduction