A Novel Reduction Algorithm Based Decomposition And Merging Strategy

In knowledge acquisition based on rough set theory, attribute reduction is a key problem. In this paper, by integrating with the idea of decomposing and merging decision tables, a novel attribute reduction algorithm with three steps is presented. First, an intact decision table is decomposed into two sub-decision tables with same condition and decision attributes, according to the decomposition strategy presented. Second, attribute reduction results of the two sub-decision tables are acquired in terms of classical attribute reduction algorithms. Third, a merging algorithm is developed, so the two sub-decision tables are merged into the intact decision table. The whole reduction process of original decision table is implemented. Experimental results demonstrate the efficiency of the proposed algorithms.