Research And Application Of Rough Set Theory In Electric Power Customer Data Mining

The innovation of electric power system begins in 1998, from then on, the electric power system of China steps into an open market gradually.In the open market of electric power system companys need to know more about their customers in order to achieve success in the competition in open market. After in-depth study of rough set theory a max confidence based algorithm of data completing named MCC is put forward,and the validity of MCC is testified by examples.A new attribute reduction algorithm based on information entropy named IEAR is put forward then ,the experiment with UCI data proves that the number of attributes left after attribute reduction can be reduced effectively by IEAR.A rough set based data mining system named RSDMS is put forward and realized after the analysis of algorithms in rough set.The mode to mining from customer data of electric power system is discussed and implemented in RSDMS.The result shows that RSDMS is effective ,and RSDMS can be used in other fields commendably.