earticle

영어

The fuzziness and randomness of decision table affect hugely on the performance of knowledge acquisition in rough set. In order to reduce their influence, a novel reduction algorithm based on grey relational analysis is proposed. In the algorithm, every value of decision table is converted to the same domain. Moreover, on the basis of grey relational analysis, the grey relational matrix for the converted decision table is constructed to describe the equivalence relations between samples of decision table. Finally, the samples with the same similar level are adopted as the coarser granularity. The experiments fully show that the reduction decision table achieved almost the same recognition rate with less than one tenth of the former conditions. It fully shows the effectiveness of the algorithm.