A Data Mining System Based On The Attribute Reduction Of Rough Set Theory

Data mining(DM) is a new hot research point in database area. Data mining gets knowledge from large quantity of data. There are some methods for data mining, and Rough Sets methodology is one of important methods. This paper researches a data mining system based on attributes reduction of rough set theory. In this paper, rough set theory has been discussed, by the analysis and synthesis of data mining algorithm based on rough set theory, two new kinds of attributes reduction algorithms which are based on genetic algorithm and heuristics algorithm, have been presented. The data mining system includes two parts, one is rules generation, it creates rules after attributes are reduced by the attributes reduction algorithms above; the other is classification, it gives a classification for new objects according to these rules which have been created in rules generation. The experimental results show that the algorithms above have been tested successfully.