A Kind Of The Decision Rules And Its Nested Structure In Decision-making Information Table

In the processing of decision information table by classic rough set theory, the partition of domain U is achieved according to the equivalence relationship which is formed by the equivalence of attribute values of elements in domain U, and the decision rules are also formed based on the equivalence relationship. In chapter 1, this dissertation mainly introduces the patition of elements in domain U by rough set theory as well as the formation and the selection of decision rules. In chapter 2, a new equivalence relationship in domain U and its corresponding decision rules with new forms are achieved through the patition to condition attribute values space , and then the dynamic relationship between the nest of condition attribute values space's patition and the nest of decision rules with new forms is researched. An example shows that the decision rules with new forms are better than the former decision rules. In chapter 3, the model based on the patition to condition attribute values space is popularized to the more general model in which the neighborhoods are defined on condition attribute values space, and then the compatible relationship in domain U and the compatible granules coverings are constructed as well as its corresponging decision rules are formed. At last the dynamic relationship between the nest of the neighborhoods defined on condition attribute values space and the nest of its decision rules with the covering model is researched as well as the connection between the two models is researched.