Heuristic algorithm for attribute reduction and its extended application

Zheng Zhang,Daoyuan Yu

2007-01-01

Abstract:The concepts of secondary core and regenerated discernibility matrix are given based on the discernibility matrix, and then the heuristic attribute reduction algorithm based on maximum discernibility degree is put forward, which reaches an effective way of gaining minimum attribute reduction of decision table. According to thinking of the hierarchical reduction approach, the hierarchical reduction approach based on the significance of attribute is put forward. The approach, which does not lead the information losing of decision system, satisfies the fast reduction requirements of dynamical system, and has better practicability. The incomplete information system, enlightened on the recognition laws of human being, is divided into complete attribute level and incomplete attribute level. Firstly, complete attribute level can be reduced by the heuristic attribute reduction algorithm based on maximum discernibility degree. Secondly, the incomplete attribute can be reduced according to its intactness degree one by one. Lastly, the rules can be extracted rightly. The reduction approach can decrease the uncertainty because of completeness process. The practical results show this approach both are fit for the attribute reduce or rules acquisition of complete and incomplete information system.