The Research And Application Of DM Algorithm On The Basis Of Rough Set

With the rapid increase of the data quantity, data mining has become an important means to discover knowledge from a large number of data. Rough Set is a tool to deal with vague and uncertain data, therefore it becomes an important frame in DM. Reduction and core are the key content of Rough Set theory. Data after reduction is more valuable and can obtain knowledge more accurately.First of all, on basis of the results of the academic research on both domestic and international relevant DM and KDD, the methods of data discretization and data generalization, the basic algorithm of attribute reduction, and the typical algorithm of attribute reduction are discussed in this thesis, which includes the algorithm based on logic operation , frequency , attribute dependence and entropy. The time complexity of each reduction algorithm is analyzed. In addition, the reduction algorithm of deleting the rear attribute is proposed which does not need core calculating and can save time and space and simplify the solving process as well.Secondly, based on attribute reduction, the algorithm of drawing the simplest and typical rules and building decision tree are studied in this thesis. Also the algorithm of how to build decision tree based on Rough Set is proposed and its correctness with examples based on the discussion of ID3 algorithm is verified.Finally, on the basis of the theory methods research mentioned above,the tool of KDD based on Rough Set is designed in this thesis.