Study On Attribute Reduction Algorithm Based On Rough Set And Fuzzy Set Theory

The research about attributes reduction is always the key problem of data mining domain and the key technique for an enterprise's decision knowledge acquisition.However,the attributes reduction algorithm existing still call for deeper study whatever in the algorithm's effeciency and storing capacity or the appling range.At once,a breakthrough is achieved in the algorithm research,it will dirve forward data mining technology used in practice strenuously,and increase enterpise's decision efficiency too.The paper researched and analyzed the existing research results of knowledge acquisition base on rough set and fuzzy set theory,and find that it will be more effective on attributes reduction when they are combined. By in-depth analyzing decision table attributes reduction base on rough set and fuzzy set theory,a more effective reduction strategy is presented. By combining heuristic attribute reduction algorithm with attributes reduction method based on discernibility matrix,this paper present an improvement algorithm,which can cut down the original algorithm's computing and storing capacity, and make it apply discrete decision tables too.The paper also researeched recuction algorithm about continuous attributes decision table,gave an improved algorithm for a quickly attribute reduction algorithm,decreasing its computing times and enhancing its reliability. By a large number of experimental data with UCI data sets,further explains the improved algorithm is more effective.