Financial Risks Classification Early Warning Analysis of Data Mining Technology

Abstract

To further improve the informatization level of financial risk early warning, a financial risk classification early warning method based on neural network quantile regression algorithm is proposed. Among them, macro and micro indicators are selected as the index input of early warning, and then the neural network quantile regression algorithm is used to classify and warn the financial risks, finally the specific risk level is output. Simulation results show that the MAE and RMSE of neural network quantile regression algorithm are 7.12e-09 and 1.301e-08, which are lower than those of BP neural network and generalized neural network. Thus the superiority of neural network quantile regression model is verified.