Dynamic Maintenance of Decision Rules with Rough Set Under Characteristic Relation

Rough sets for knowledge update have been successfully applied in data mining. Methods for incremental updating decision rules based on the indiscernibility, tolerance relation and similarity relations in rough set theory have been previously studied in literature. The characteristic relation-based rough sets approach provides more informative results than the approach employing the indiscernibility, tolerance relations and similarity relations based approach. In this paper, we extend rough sets based on characteristic relations for incrementally updating decision rules. An extensive experimental evaluation validates the efficiency of the proposed approach which may be used to handle a dynamic maintenance of decision rules in data mining.