基于灰色预测–决策树模型下车险保费浮动的研究  

Study on the Floating Premium of Car Insurance Based on Grey Prediction Method and Decision Tree

作　　者：肖思佳 林萍芝 张靖文 王浩华[1] 

机构地区：[1]海南大学理学院,海南 海口

出　　处：《应用数学进展》2021年第7期2579-2591,共13页Advances in Applied Mathematics

摘　　要：汽车保险的续保问题对于保险市场具有举足轻重的作用,如何有效的预测市场各个因素对续保率的影响并未十分清楚。通过分析保费浮动优惠等因素对续保率的影响,本文建立影响力指数的等级比较模型,通过计算其性质系数的灰色关联度,考察具有不同性质车辆的续保率。计算出有关索赔系数和使用性质系数,建立数学模型,给出了较为合理的保费浮动方案。最后,建立决策树模型,根据统计数据对不同车辆进行分类同时,通过对三个因素进行加权,得到了评价企业质量的决策树累积模型,分析各因素对续保率的影响,从而提供可行性建议。This paper intends to predict the floating premium of car insurance in an insurance company, use the renewal rate to measure the floating demand of premium, collect the data of relevant departments, and establish the grade comparison model of influence index by analyzing the influence of each factor on the renewal rate. Then the grey correlation degree is used to predict the influence of vehicles with different use properties on the renewal rate, and the use property coefficient is calculated. According to the model of “premium floating coefficient = claim coefficient A ×use property coefficient B”, a more reasonable premium floating scheme is put forward. Finally, the decision tree is established, combined with the statistical data, the vehicles with different usage properties are weighted, and the three factors that evaluate the quality of the company are assigned weights, and the cumulative model of the decision tree is obtained. The results show that the scheme evaluation is effective.

关 键 词：等级比较模型 续保率 灰色预测法 保费浮动系数 决策树 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术]