Rough set attribute reduction based on mutual information and Beam search

The attribute reduction problem for rough set is analyzed by the mutual information of attribute set. Based on mutual information, the redundancy-synergy coefficient of attribute set, a novel measure for redundancy and synergistic ability of attribute set, is defined. A Beam search based attribute reduction algorithm for rough set is presented, where the redundancy-synergy coefficient is taken as the attribute reduction measure. Experiments show that the new algorithm yields satisfying attribute reduction results