Fuzziness-Preserving Attribute Reduction from Hybrid Data

In this paper, we devote to present a fuzziness-preserving attribute reduction in fuzzy rough set framework. Through constructing the membership function of an object, we first introduce a fuzzy measure to assess the fuzziness of a fuzzy rough set and a fuzzy rough decision, which underlies a foundation for attribute reduction algorithm. Then, we derive an attribute significance measure based on the proposed fuzzy measure and design a forward greedy algorithm (ARBF) for attribute reduction from hybrid. Numerical experiments show the validity of the proposed algorithm from search strategy and heuristic function in the meaning of fuzziness-preserving.