Research On Approaches Of Rough Set Data Analysis

Knowledge discovery is a tool for mining knowledge from data, and it concentrates on obtaining the effective, novel, potential useful and understandable pattern. Rough set theory was proposed by Z. Pawlak In 1982, in which those individuals that can't be determined belong to border area, base on which reduction algorithm of information table and decision table was proposed. But, the traditional rough set theory still have several problems need to solving. For example, it is still a NP-hand question in finding all reduction. Rough set theory has been successfully used in intelligent artificial and recognize area as a processing method of uncertain data.The thesis first introduced the meaning and basic process of knowledge discovery, and then presented the basic theory and method of rough set. The main work of the paper is as follows.In the aspect of the method of computing technology of rough set . The matrix algorithm under the information system is proposed by W. Guan, D. A. Bell and Z. Guan, in which the equivalent relation under the information system was described in the form of matrix, and based on which a new reduction algorithm was proposed. In this basis, the concept of positive vector set was proposed, and the importance of every attribute was measured by the changing of positive vector set in the process of relative reduction. And then a reduction algorithm was given for attribute in decision system. Then, tolerance relation was described by matrix, and the reduction algorithm of matrix in complete information table was extended to incomplete information table, which make the matrix algorithm have further application.On the other hand, in some cases, some attribute unable receive certain value in decision system, but its general range sometimes can be described by a area or some possible values. The traditional rough set theory is unable...