Dynamic Update Methods Of Multi-granulation Rough Sets

With the development of information technology and the continuous update of information systems,processing complex and huge dynamic data is facing serious computational inefficiency.Dynamic calculation provides an effective way to solve this problem.In a complete information system,most approximate dynamic update methods are mainly for the classic rough set and its extended model.In a multi-granulation environment,the current research methods are few and the time efficiency is not high.In addition,in many practical applications,the data that needs to be processed is not a single one,but a complex data that is a mixture of numerical and symbolic types.At the same time,information systems are often not necessarily incomplete information systems(that is,data will be missing or non-existent).At present,there are few researches on the approximate dynamic update of information system with incomplete mixed data.Therefore,based on the complete and incomplete information systems,this paper studies the dynamic update issues separately,mainly including the following:(1)For the multi-granulation rough sets in the complete information system,when the attribute values in the information system change(the attribute values are coarsened or refined),the relative relationship matrix is obtained by reducing the calculation of objects in the universe that have nothing to do with the target concept.The approximate sets of the multi-granulation rough sets are represented by the matrix operation of Boolean matrix,cut matrix and relative relation matrix.The change trend of the approximate sets when the attribute values are coarsened or refined is studied,and the method of dynamic update of the approximate sets based on the relative relationship matrix is further proposed and the corresponding efficient algorithm is designed.Finally,the feasibility and effectiveness of the algorithm are verified through the UCI data sets.(2)For the incomplete mixed data information system,multi-granulation rough sets model based on the neighborhood tolerance relationship is proposed,and its basic properties are discussed.In addition,the change of the approximate sets of the neighborhood multi-granulation rough sets is studied when the missing attribute values in the system are obtained one by one.According to the changing trend of the approximate set,the dynamic update theorems of the neighborhood tolerance class are proposed,and the dynamic update method of the approximate sets of the neighborhood multi-granulation rough set is obtained.Finally,an example is used to verify the feasibility of the proposed method.