Data Classification Modeling Based on Granular Computing

Problem of data classification was theoretically studied based on Granular Computing(GrC).Concept of AllGS(All-Granular-Space) was first introduced,and then with GrC a set of data was expressed in granulation,and concept learning was rationally explained,which lead to analysis of mechanism of data classification. At last,KDD modeling of data classification was educed,with which a theoretical basis was given to facing-problem-solving of KDD,and a solid foundation was also laid for further research.