Research On Construction And Knowledge Updating Of Multi-granulation Rough Sets Model

This paper combines multi-granulation rough sets with covering rough sets,studies the multi-granulation rough set model based on covering in the sense of the covering and granularity.Under the background of incom plete multi-granulation rough sets,this paper studies the dynamic knowledge update and puts forward effective algorithms to obtain knowledge.Firstly,this paper gives the method of constructing a multi-granulation rough set model from a covering rough set,studies the relevant properties of the model,and discusses the relationship s between the upper and lower approximations and reductions of loose covering rough sets and optimistic multi-granulation rough sets,strong covering rough sets and pessimistic multi-granulation rough sets.Further,under the two different conditions of deleting the intersection reducible elements or union reducible elements in a multigranulation approximation space and a covering approximation space,using the given construction method,two different multi-granulation rough sets based on covering are obtained.Finally,combined with an example,the relationships between the two multigranulation rough sets based on covering and the original multi-granulation rough sets are compared.In the incomplete information system with missing attribute values,based on the tolerance relationship,this paper gives the definitions of lower and upper approximations of optimistic multi-granulation rough sets and pessimistic multi-granulation rough sets when adding and deleting objects in the universe.In addition,this paper studies the related properties,designs two dynamic algorithms for adding and deleting objects,and analyzes the time complexity of the two algorithms.Further,three groups of experiments are designed from two aspects of data scale and data update rate,the effectiveness of the proposed dynamic algorithms is verified.