基于区分矩阵的最小规则挖掘算法

An Algorithm of Minimum Rule Mining Based on Discernable Matrixes

从区分矩阵的概念出发,分析了区分矩阵与属性约简的关系,并在此基础上设计了一种基于对区分矩阵进行缩减的规则挖掘算法.该算法使得规则提取过程相对清晰和简单,当数据量很大时,该算法降低了计算的复杂性.通过实例验证了该算法的有效性.

From the conception of discernable matrixes, the relation between discernable matrixes and attribute reduce is analyzed in this paper. Based on the analysis about the discernable matrixes, we presented an algorithm of minimum rule mining based on reducing discernable matrixes. This algorithm makes the attribute reducing explicit and reduces the complexity of computation when the volume of data is large. At last , the validity is proved using an example.