一种基于零值原则的属性约简方法  

Attribute reduction method based on zero value principle

作　　者：罗来鹏[1] 刘二根[1] 

机构地区：[1]华东交通大学基础科学学院,南昌330013

出　　处：《计算机应用》2008年第4期872-873,876,共3页journal of Computer Applications

基　　金：国家自然科学基金资助项目(10661007);江西省自然科学基金资助项目(0611009);华东交通大学校立科研基金资助项目(07JC05)

摘　　要：根据Guan等提出的完备信息系统下矩阵约简算法,提出一种改进的属性约简计算方法。该方法根据矩阵的运算特点,通过引入唯一零值概念,使得计算过程更为简易。证明了它与区分矩阵下属性约简的等价性,最后将该方法运用到协调决策表中,并用实例对此进行了说明。Based on matrix reduction algorithm for information system introduced by Guan, a new improved computational method of attribute reduction was presented. According to the characteristics of operation between matrixes, the computation was simplified by introducing only the concept of zero value. This method was proved to be equal to attribute reduction by discernable matrix. At last, the method was applied to complete decision table. Its correctness and effectiveness are shown in an example.

关 键 词：粗糙集 属性矩阵 属性约简 

分 类 号：TP311.32[自动化与计算机技术—计算机软件与理论]