A New Approach to Attribute Reduction of Decision Information Systems

Rough set theory is a theoretical tool for complete conclusions by incomplete information; its basic idea is using some partition to consider the description problem of concept on university. In this paper, we firstly introduce some usual attribute reduction methods. Secondly, we establish an effect-based roughness measure model, named as effect rough degree (ERD) by combining with a basic measure factor for roughness-lower (upper) accuracy of rough set. Finally, we propose an ERD-based attribute reduction method (abbreviated as ERD-RM) and then discuss the difference and relation between ERD-RM and the existing reduction methods. The theoretical analysis showed that the ERD-RM includes other reduction methods to a certain degree, and it also has basic difference with other existing methods.