Application of Improved PSO-ELM in Auto Insurance Customer Risk Level Prediction
Xi Chen, Liang Xu, Yalong Wang, Xiang Zhai, Xiang Guo · 2020
In order to improve the prediction accuracy of automobile insurance customer’s risk level, a prediction model of automobile insurance customer’s risk level based on GSPSO - ELM is proposed. The premium, insurance amount, number of risks, gender, driving age and other indicators were used as the main criteria to predict the risk level. The simulation results show that the proposed GSPSO - ELM prediction model can not only overcome the local optimal, and has higher precision of prediction, forecasting results of the mean square error (mse), decision coefficient and mean absolute percentage error and dynamic inertia weight PSO algorithm to optimize extreme learning machine, linear decreasing inertia weight PSO algorithm to optimize extreme learning machine model are improved.