In order to process the incomplete decision system with hybrid value, a classification method based on the double-neighborhood rough set model is proposed in this paper. Firstly, an uncertain distance function --- Connection degree distance function is defined, and the double-neighborhood rough set model based on connection degree distance function is constructed. Then, based on the constructed model, an attribute reduction algorithm is discussed, and a classification algorithm based on attribute reduction and covering reduction is provided. Finally, some experiments are carried out on UCI data sets. The experiments results show that the proposed classification algorithm is objective and effective, and it is more effective when the missing value is abounded.