A Detection Method for Electricity Theft Based on Random Forest Algorithm

The electricity theft behavior is becoming more and more common in order to reduce the cost of consumption. A detection method for electricity theft based on random forest is proposed in this paper. Besides, the evaluation index system including the power consumption slope index, the line-loss index, and the warning information index is built to consider the factors affecting the electricity theft. For improving the detection accuracy, the data cleaning and miss value preprocess are conducted before training the decision trees. Finally, a numerical case and the comparison in performance with decision tree and GBDT are used to verify the availability and superiority of the proposed detection method.