A rule extraction algorithm based on compound attribute measure in decision systems

With introduction of information granularity in decision systems in this paper, the importance of core attributes and information granularity is analyzed. Besides, an effective compound attribute measure is defined, which not only considers the measures of certain information in the positive region, but also considers the importance of information granularity beyond the positive region. Based on the proposed compound attribute measure, an efficient rule extraction algorithm is presented in decision systems. Before mining the classification rules, the redundant attributes are removed in the attribute reduction stage, such that the algorithm can extract brief classification rules. Finally, a case study further verifies the feasibility and efficiency of the proposed algorithm.