Abstract:

To derive the positive and negative decision rules from incomplete noise information systems, the concept of variable precision rough set descriptors is proposed, and its properties are discussed. Then, the approach to acquire the two kinds of decision rules by the proposed variable precision rough set descriptors is presented. To obtain the simplified rules, an approach of discernibility matrix is proposed to obtain the reduction of the descriptors, and a heuristic algorithm of attribute reduction is presented, which can preserve the lower approximate of the descriptor. Finally, some examples of student globe evaluation show the effectiveness of the approach.