An Attribute Reduction Algorithm by Switching Exhaustive and Heuristic Computation of Relative Reducts
Yasuo Kudo, Tetsuya Murai · 2010
We propose a heuristic algorithm to compute as many relative reducts as possible from a decision table with numerous condition attributes. The proposed algorithm is based on generating many reduced decision tables that preserve discernibility of objects in the given decision table. Moreover, the proposed algorithm switches exhaustive attribute reduction and heuristic attribute reduction by the number of condition attributes in decision tables. Experimental results indicate that the proposed algorithm can generate many relative reducts from datasets that are difficult to compute all relative reducts.