All Times ET
Keywords: Claim Data, Health Insurance, Statistical Distribution
This research focuses on the description of a method of analyzing insurance claim data in order to determine the distribution and pattern of the claims. The specific objective is to study the claim behavior of a certain health insurance product from past experiences and to apply the most appropriate methods that fit the available data. This work solved two major problems faced by general insurance companies when they want to use past or present claim amounts in forecasting future claims. Mainly, they must find an appropriate statistical distribution for a large volume of claim amounts and test how well this statistical distribution fits their claim data. This project sets forth a methodology for dealing with these problems. In this study, an exploratory analysis of the claim data was carried out to visualize the shape of the data. Log-likelihood was computed for five probability families in order to select the best probability distribution. The Goodness of fit test was carried out using Q-Q plots and Akaike’s Information Criterion (AIC) values were computed. The result shows that the health claim data follows a lognormal distribution. Due to the skewness of the data, a log-transformation was carried out. The log-transformed data follows a normal distribution using the log-likelihood and AIC test as criteria for selection.