Prediction of default probability of clients´ electricity charge arrears

On the basis of analyzing the reasons bringing on electricity charge arrears and utilizing the gettable data, the key variables effecting on the charge arrears which are needed in arrear recognition risk are designed. By means of logistic regression theory which is commonly used in dealing with bi-category problem, we established the model which can recognize the possibility of electric power clients´ charge arrears. Thereby, it is possible to predict the default probability according to the latest data from client that the electric power enterprises could acquire, and the previous management mode is to be changed, i.e. passive occurrence of charge arrears are turned into active prevention in advance. Therefore propose of preventing client arrears could be achieved.