Attributes Reduction Based on Rough Set

Attributes reduction is one major problems in rough set theory. A method of attributes reduction based on scan vector is proposed in this paper. Firstly, define a new conception of discernible vector by which we can transform the information table into discernible vector set. Secondly, a plus rule for the discernible vector based on its good structure is defined, and consequently we can obtain a scan vector through scanning the discernible vector just only one time, which can represent the information table better because the scan vector has a more concise structure. And then, take the attribute frequency vector in the scan vector as the heuristic information and search for the attributes reduction in the discernible attributes set which has less numbers of elements than the original. Finally, the experiments results indicate that the method proposed in this paper is much more effective