广义线性模型在UBI车险费率厘定中的应用

发布时间：2018-04-16 15:23

本文选题：广义线性模型 + 车险费率　；参考：《天津商业大学》2017年硕士论文

【摘要】：当前中国的汽车产业处在一个高速发展的阶段,中国的汽车销量已经连续多年位居全球第一。近年来,中国汽车保有量持续增长,成为仅次于美国的世界第二大汽车保有国[1],预计2019年将超越美国成为世界第一大汽车保有国。由于我国汽车产业和汽车保有量的高速发展,与之相对应的车险市场也呈现出快速发展的态势。然而,随着车联网和大数据等新技术的出现,新型的车险费率厘定模式在技术上得到突破,继而相应的创新也得以实现;加之,随着我国车险费率市场化改革的发展,我国现有的车险费率厘定模式已难以满足新形势的要求,加快实现车险费率差异化和个性化己经成为车险市场发展的必然趋势[2]。近年来国内外保险公司都在积极推进新型的机动车辆保险产品UBI(Usage-Based Insurance),依据驾驶行为、行驶里程等分析确定出驾驶行为风险等级,为车险投保人提供公平合理的车险费率、差异化和个性化的车险策略以及车险增值服务[3]。本文尝试采用广义线性模型对UBI车险的风险等级进行确定并且对驾驶员的驾驶行为进行评分,然后在此基础上厘定得出最终的车险费率。本文的主要工作内容如下:(1)首先对当前已有的费率厘定模式及其优缺点进行说明和分析,并着重分析了传统费率厘定的各种方法所存在的不足。然后介绍了车联网技术和车联网理论的发展情况以及应用情况,并介绍了车联网保险UBI的研究现状和发展趋势。在此基础上着重介绍了驾驶员驾驶行为对驾驶安全或者风险影响的内容和方式。(2)在车联网环境下,在对驾驶行为对风险水平的影响进行分析的基础上,采用广义线性模型,建立了基于驾驶行为数据和保单数据的风险评分模型,从而得到每种驾驶行为类型的评分结果。以此为基础,将驾驶行为评分作为新的定价因子,并与传统的从人从车定价因子相结合,建立广义线性模型进行费率厘定,最终得到折扣系数,进而计算得到保费。
[Abstract]:At present, China's auto industry is in a stage of rapid development, China's auto sales have been ranked first in the world for many years.In recent years, China's auto ownership has continued to grow, becoming the world's second largest car holder after the United States, and is expected to overtake the United States as the world's largest car holder in 2019.Due to the rapid development of automobile industry and vehicle ownership in China, the corresponding auto insurance market is also showing a rapid development trend.However, with the emergence of new technologies such as car networking and big data, the new model of vehicle insurance rate determination has made a breakthrough in technology, and the corresponding innovation has also been realized; in addition, with the development of market-oriented reform of car insurance rate in China,It is difficult to meet the requirements of the new situation in the existing car insurance rate determination model in our country. Speeding up the realization of the difference and individualization of vehicle insurance rates has become an inevitable trend in the development of the auto insurance market [2].In recent years, insurance companies at home and abroad are actively promoting the new type of motor vehicle insurance product UBI(Usage-Based Insurance, according to driving behavior, driving mileage and other analysis to determine the risk of driving behavior, to provide a fair and reasonable car insurance rate for vehicle insurance policy holders.Differentiated and personalized auto insurance strategy and auto insurance value-added service [3].This paper attempts to use the generalized linear model to determine the risk level of UBI vehicle insurance and to score the driver's driving behavior, and then determine the final vehicle insurance rate on this basis.The main work of this paper is as follows: (1) first of all, this paper explains and analyzes the existing models of rate determination and their advantages and disadvantages, and emphatically analyzes the shortcomings of the traditional methods of rate determination.Then it introduces the development and application of vehicle networking technology and vehicle networking theory, and introduces the research status and development trend of vehicle networking insurance UBI.On this basis, the content and method of driver's driving behavior influence on driving safety or risk are introduced emphatically. In the environment of vehicle network, the influence of driving behavior on risk level is analyzed, and then the generalized linear model is adopted.A risk rating model based on driving behavior data and policy data is established, and the results of each driving behavior type are obtained.On this basis, the driving behavior score is taken as a new pricing factor, and combined with the traditional slave car pricing factor, a generalized linear model is established to determine the rate, and finally the discount coefficient is obtained, and then the premium is calculated.
【学位授予单位】：天津商业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：F842.634

【参考文献】