Research On The Evaluation Of Outstanding Claims Based On The Improved Multiple Interpolation Method

Data loss is an unavoidable phenomenon in insurance actuarial calculations.Once the missing data appears without proper treatment,it will greatly affect the estimation results of the reserves,thus increasing the operational risk of insurance companies.It is an important research content of actuarial and statistical analysis of non-life insurance to deal with the missing value reasonably.This paper firstly introduces the theoretical knowledge of missing data.According to the missing data interpolation criterion,the single interpolation method and multiple interpolation method are compared and analyzed.According to the characteristic of upper flow triangle data distribution of insurance history claim data,the multiple interpolation method is improved,so that the improved multiple interpolation method can more accurately interpolate missing data.Then,based on the method of estimation of unearned indemnity reserves and relevant basic theoretical knowledge,the commonly used models and methods of the evaluation of unearned indemnity reserves are summarized:chain ladder method,reserve progress method and generalized linear model.At the same time,the paper makes an in-depth analysis and research on the related issues in the evaluation process of the pending indemnity reserve.Finally,the paper makes an empirical analysis of various missing data interpolation methods and reserve evaluation models involved in the previous chapters.The main work includes two points.First,through R language programming,the missing values in the flow triangle are treated by mean interpolation,random interpolation,multiple interpolation and improved multiple interpolation,and the interpolation results are compared and analyzed.Second,lack of the above five kinds of data processing method is applied to the separately applied to the chain ladder method,reserve progress method,generalized linear model of three kinds of evaluating the outstanding claims reserve evaluation model,the related estimation results and conclusions are obtained:The improved multiple interpolation method can better interpolate the missing value in the insurance data and obtain more accurate estimation results of the pending indemnity reserve.