Research On Auto Insurance Rate Determination Based On Big Data

With the implementation of the market-oriented reform of commercial auto insurance premium rate and the application of big data and other technologies in insurance premium rate pricing,the traditional auto insurance premium rate only considers the risk factor of "subordinate car" and the auto insurance products of various insurance companies are similar to each other,which has completely failed to adapt to the differentiated and personalized pricing demands of the insurance industry.If property insurance companies want to adapt to the marketization of auto insurance rate and supply auto insurance products that meet the needs of consumers,they must take more account of the driving environment and drivers’ own risk factors to determine the impact factor of the rate,and give full play to the advantages of big data in calculating the rate.This thesis mainly USES the methods of non-parametric chi-square test and zeroexpansion poisson regression model to study the data of traffic flow at an intersection and traffic accident in a provincial capital city provided by the platform of Open ITS alliance,and analyzes the traffic flow in the driving environment,driving time in the morning and evening peak,climate,precipitation and other risk factors.Explore the driver’s own risk factors.The study shows that the risk factors in driving environment have significant influence on traffic accidents.The difference of gender,age and driving age is related to the risk of automobile insurance,and a more accurate reference for determining the premium rate of automobile insurance is put forward: the driving environment and the driver’s risk factor are taken into account to provide a new pricing direction for property insurance companies.The main innovations of this thesis are as follows: the innovation in the topic selection and research ideas,which have strong practical value and certain frontier in the topic selection,and the research ideas,further verify the driving age factor of drivers’ risk factors.The disadvantages are as follows: on the basis of referential literature,the pricing analysis of the risk factors to determine the rates needs to be further conducted by the model,but this involves more private data of the policyholders,so it is impossible to accurately verify the auto insurance pricing.