Research On Algorithms Of Rule Reduction And Attribute Reduction

More and more people pay attention to data mining (DM) as the application of database system and the accumulation of plenty data. The application and the theory of rough set have a great success, especially in the area of machine learning, knowledge acquisition, intelligent control system, decision analysis, expert system and pattern recognition.In this thesis, a detail study of methods and processes of data mining with the tool of rough set theory is carried out. The problem of knowledge discovery, rule optimization and reduction is studied based on the basic theory of rough set.The main research is as follows:In chapter 1, the meaning and the aim of the research and the status in quo inland and overseas are represented.In chapter 2, firstly the DM is introduced, and the basic theory of the rough set is studied, in which includes the approximate set and rough set, attribute reduction and dependence, the characteristic of rough set theory, the application of rough set theory in data mining. Then the information system and the decision table are introduced, in which includes the basic concept of rough set and the discrimination function, decision rules and the application rough set in the information system and decision table.In chapter 3, the discovery algotithms of the certain rule reduction set is studied. Firstly, some interrelated concepts are introduced, and then an approach of finding out certain rule reduction set is given. It has a reliable basic theoretics.In chapter 4, the attribute reduction and rule optimization of incomplete information system are studied. The description reduction, GS reduction and DS reduction are brought forward, and they are the general condition of the approximate reduction.In chapter 5, the simplified discrimination function without core and the approach of computing are studied; an attribute reduction based on the simplified discrimination function without relative core is given, by which the problems of the expenditure of time and space in calculating the discrimination function are figured out.In chapter 6, the study work is sumed up and task that we will do in the future is put forward.