Rough set model based on credibility measure

Probabilistic rough set model is based on probabilistic measure which satisfied additive property. But in practical applications, there exist some non-additive set functions. So combining with credibility measure with self-duality and non-additive property, the conditional credibility measure is introduced, the entire credibility formula is given, the rough set model based on credibility measure is constructed, some properties of this model is proved. This model is applied to the Bayesian decision. Finally, the difference between this model and Pawlak rough set model is discussed.