基于改进XGBoost算法的UBI车险费率等级判定模型研究  

RESEARCH ON UBI VEHICLE INSURANCE RATING MODEL BASED ON IMPROVED XGBOOST ALGORITHM

作　　者：闫春 荆建业 孙晓红 刘新红[3] Yan Chun;Jing Jianye;Sun Xiaohong;Liu Xinhong(School of Mathematics and Systems Science,Shandong University of Science and Technology,Qingdao 266590,Shandong,China;Department of Audit,Shandong University of Science and Technology,Qingdao 266590,Shandong,China;Department of Mathematics and Physics,Beijing Institute of Petrochemical Technology,Beijing 102617,China)

机构地区：[1]山东科技大学数学与系统科学学院,山东青岛266590 [2]山东科技大学审计处,山东青岛266590 [3]北京石油化工学院数理系,北京102617

出　　处：《计算机应用与软件》2022年第10期254-258,349,共6页Computer Applications and Software

基　　金：国家自然科学基金项目(61502280);全国统计科学研究重点项目(2019LZ10);北京市教育委员会科技计划一般项目(KM202010017001)。

摘　　要：针对XGBoost算法参数多且不易确定从而影响算法分类预测准确度的问题,提出GA-XGBoost算法,建立UBI车险费率等级评定模型。该模型是利用遗传算法对XGBoost模型参数进行寻优,提高XGBoost模型判定的准确度。实证结果表明GA-XGBoost算法相比于SVM算法、CART算法在UBI客户驾驶行为的风险等级评定过程中具有更高的判别精度,且在建立UBI车险费率等级判定模型过程中易于实现,具有良好的鲁棒性,能够取得较理想的等级评定效果。Aimed at the problem that XGBoost algorithm’s parameters are numerous and difficult to determine,which affects the accuracy of algorithm classification and prediction,GA-XGBoost algorithm is proposed and UBI vehicle insurance rate rating model is established.The model used genetic algorithm to optimize the parameters of XGBoost model and improved the accuracy of XGBoost model.The empirical results show that GA-XGBoost algorithm has higher accuracy than SVM algorithm and CART algorithm in the process of risk rating of driving behavior of UBI customers,and it is easy to realize in the process of establishing UBI vehicle insurance rate rating model,with good robustness,and can get a ideal rating effect.

关 键 词：驾驶行为保险 XGBoost算法 遗传算法 费率等级判定 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]