Research Of Attributes Reduction And Application Based On Rough Set Theory

The rough set theory is a mathematics tool in processing inaccurate, inconsistent and incomplete problems, which can find the implicit knowledge and potential regulations by directly analyzing and deducing the data without any prior information except the data set. Since the end of 1980s, the theory and applications of rough set gradually have become the focus of intellectual information processing, and used in the fields of data mining, machine learning, pattern recongnition and decision analysis, etc.The attribute reduction is a important problem of the rough set theory , which delete the redudant attribtes on the condition of keeping on the invariable classifying ability and the fast algorithm of reduction is one of the main research contents of the theory of rough set, which is a key step of knowledge acquisition. Therefore, attribute reduction is focused on in this dissertation. The main content is divided as follows.(1) Relevant knowledge of the rough set theory is introduced and some classic attribute reduction algorithms are analyzed systematically(2) An improved algorithm of attribute reduction based on discernibility matrix IAARBDM(Improved Algorithm of Attribute Reduction Based on Discernibility Matrix) is proposed, which reduces the redundant elements by a judugement machanism before the structure of one-dimensional array is designed for storing the elements of the matrix, simplifies the discernibility function and enhances the efficiency. Furtherly an algorithm of AARSDM(Acquiring Attribute Reduction from the Simplified Discernibility Matrix) based on IAARBDM algorithm is proposed for the complex problems of disernibility function in larger database, which is useful for obtaining optimum or relatively optimum reductions.(3) A kind of heuristic fast algorithm for attribute reduction algorithm HAARDT(Heuristic Algorithm of Attribute Reduction for Decision Table) is given, which utilizes the matrix elements with the same cardinal number and the attribute frequency to verify the importance of attribute, finds out the optimum or relatively optimum reduction and reduces the calculation complexity of solving analysis normal of discriminate function.(4) An algorithm of AAAR(acquiring of the abnormal action regulations) for mining the abnormal action regulations is introduced based on rough set theory and the application value of mining abnormal action regulations is discussed by examples.