一种基于粗集理论的动态近似规则挖掘推理方法

提出一种基于粗集理论的,把属性的重要性和属性值的出现频率综合起来进行规则推理的方法.分析了“激活一个,否则离开”原则的优缺点,指出在近似推理中,大前提中的规则数量应该可变.给出一种根据推理过程中规则的出现频率决定其是否保留,从而实现规则数量的动态变化的方法,证明了动态变化过程中规则的数量不会无限增加.实例表明此法是比较有效的.

Dynamics approximate rule mining inference approach based on rough set theory

Based on rough set theory, a rule inference approach of integrating the importance of attributes with the emergence frequency of attribute value is proposed. After the analysis of the principle of "fall one or leave", it is pointed out that the number of rules in rule inference would be variable. A method is suggested to decide whether a rule is needed or not by its emergence frequency in the inference process, so the number of the rules is dynamic. The dynamic rule is proved not to increase in the process. The results of examples show the efficiency of the approach.