A Non-Parametric Method of Electric Power Enterprise Arrears Prediction Model

Power supply enterprises are facing power customer promises to bring the risk of arrears, in order to avoid the risk of customer arrears, a large power customer arrear predicting model is of great significance, through analysis of large customers to quarter data to predict the coming development trend of arrears to guide the power supply enterprise to make decisions. In response to this demand put forward a nonparametric method to predict the trend of arrears. The method by comparing a fundraising events of recent trends in activity and collected plenty of historical data on electricity consumption and dynamic data trend of delinquent trend was predicted. As the power supply enterprise to develop recovery plans based on electricity, it establishes risk control principles and strategies. The experimental results show that the algorithm maintains a low error rate, which verifies the effectiveness of the algorithm.