基于GA-BP神经网络的车险索赔频率预测与优化研究  

Research on Prediction and Optimization ofVehicle Insurance Claim Frequency Based on GA-BP Neural Network

作　　者：肖阳田 肖鸿民 XIAO Yangtian;XIAO Hongmin(College of Mathematics and Statistics,Northwest Normal University,Lanzhou 730070,China)

机构地区：[1]西北师范大学数学与统计学院,兰州730070

出　　处：《长春工程学院学报（自然科学版）》2024年第1期78-84,共7页Journal of Changchun Institute of Technology：Natural Sciences Edition

基　　金：国家自然科学基金项目(12061066);甘肃省自然科学基金(20JR5RA528)。

摘　　要：车险索赔频率的预测对于车险定价有着重要意义,近些年来,随着大数据技术的兴起,传统车险定价模型已经不能满足现在保险公司维度越来越高的大量的客户数据需求,为了提升车险索赔频率的预测精度,采用法国某保险公司车险客户的真实数据,将遗传算法加入到BP神经网络之中,对相关模型进行比较来选择最优模型。研究结果表明:遗传算法优化模型的预测精度明显优于BP神经网络,对车险索赔频率预测性能更好,可以有效降低车险定价成本。The prediction of the frequency of car insurance claims is of great significance for car insurance pricing.In recent years,with the rise of big data technology,traditional car insurance pricing models can no longer meet the increasing demand for a large amount of customer data from insurance companies.In order to improve the prediction accuracy of car insurance claim frequency,a real data of car insurance customers from a French insurance company is used,and a genetic algorithm is added to the BP neural network to compare the relevant models and select the optimal model.The research results indicate that the prediction accuracy of the genetic algorithm optimization model is significantly better than that of the BP neural network,and its performance in predicting the frequency of car insurance claims is better,which can effectively reduce the pricing cost of car insurance.

关 键 词：汽车保险 索赔频率 遗传算法 BP神经网络 ROC曲线 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]