Research On Life Insurance Underwriting Risk Based On Big Data Technology

With the deep-seated development of China’s insurance industry,the risk awareness of ordinary consumers has been constantly improved,and their understanding of insurance has been deepened.The national level has vigorously promoted the continuous implementation of commercial endowment insurance,commercial health insurance,and major illness insurance,and the insurance industry has ushered in tremendous development opportunities.Underwriting plays an important role in the whole life cycle of insurance as an important link for insurance companies to identify the risks of customers.How to improve the efficiency of risk identification,effectively find out the adverse selection of the insured in the insurance process and insurance with diseases,and reduce the company’s compensation risk has always been an important work for underwriting workers.With the deep popularization of mobile Internet,various wearable devices are also accepted by the majority of consumers.Customers have generated a large amount of data in the Internet space.At the same time,with the continuous development of big data technology and artificial intelligence,as well as the natural thirst of insurance companies for big data,insurance companies are actively trying in precision marketing,and accurate pricing,prevention of insurance fraud and compensation fraud.This paper focuses on the research of using big data technology to clean and process the data collected in the insurance process based on the data of the role of the applicant,the insured,the beneficiary,etc.obtained by the insurance company in the underwriting process,as well as the relevant institutional attributes of the salesperson.Finally,training data is formed.Logistic,decision tree,linear discriminant analysis and other models are used to analyze and train the underwriting model of the factors that affect underwriting judgment,Finally,an automatic underwriting prediction model is formed,and the existing underwriting process is optimized based on the model.Finally,combined with the current industry development and the regulatory dynamics of the regulators,the ideas and directions for the subsequent optimization of underwriting models are proposed.