在不完备信息系统中基于VPRSM的约简异常分析

Analysis of abnormal reduction based on VPRSM in incomplete information systems

属性约简问题的关键在于约简集合能否表达与原属性集完全一致的信息。分析了完备信息系统中基于变精度粗糙集模型属性约简异常出现的原因,并将其扩展到不完备信息系统中,在文献[1]的基础上,给出了不完备信息系统中基于集对分析的VPRSM的约简定义,既能保证在不完备信息系统中基于变精度约简的准确性,又能增加灵活性和容噪能力。

key of attributes reduction is whether reduction set can have the completely same information of original attribute set. This article analyzes the reasons of attribute reduction anomaly based on variable precision rough model in the complete information systems,expands it to incomplete information systems,and shows the simple defination tawards reduction based on VPRSM of set pair analysis in incomplete information systems.This definition not only can ensure the reduction accuracy based on VPRSM in incomplete information systems,but also can enhance its flexibility and noise capacity ability.