Attribute Reduction In Utility-Based Decision-Theoretic Rough Set Models

Decision-theoretic rough set (DTRS) model, proposed by Yao in the early 1990's, introduces Bayesian decision procedure and loss function in rough set theory. Considering utility function in decision processing, utility-based decision-theoretic rough set model (UDTRS) is given in this paper. The utility of the positive region, the boundary region and the negative region are obtained respectively. We provide a reduction definition which can obtain the maximal utility in decisions. A heuristic reduction algorithm with respect to the definition is proposed. Finally, experimental results show the proposed algorithm is effective.