False Invoicing Feature Identification and Risk Prediction

With the rapid development of economy, the behavior of false invoicing by enterprises disturbs the tax order and even harms the national interests, which has become a hot issue of social concern. Tax authorities can crack down on enterprises' false invoicing according to risk characteristics. In this paper, we analyze these behavioral characteristics. By comparing the prediction performance of each algorithm e.g. Logistic Regression, Support Vector Machine, Decision Tree, BP neural network, Random Forest, Gradient Boosting Decision Tree and GBoost classification that based on historical case data, we select that the Random Forest which is the one with the highest prediction accuracy as the final model. We use extended data to deal with complex features, and propose a high-precision prediction model based on Random Forest, which is more intelligent and efficient than traditional ones, so as to provide accurate decision-making basis for the prediction of enterprises false invoicing.