Modeling Number of Third-party Legal Liability Insurance Claims

General insurance is insurance that bears financial losses due to the destruction of some or all of the property of the insured. In general insurance, there are 13 types of insurance products that are commonly marketed, one of which is motor vehicle insurance. Risks guaranteed in motor vehicle insurance include, among others, loss and damage to vehicles covered by accident, fire, theft, and third-party liability. In third-party liability insurance, the owner of the vehicle bears the claim of another driver, passenger or pedestrian, resulting from an accident involving the owner of the insured vehicle. In insurance, there is a claim term. The occurrence of this claim may occur at any time so that the claim is a random variable. Since claim data is a count data so that ordinary regression that states there must be a linear relationship between the free and dependent variables, the error must be normally distributed, and so forth cannot be used. So, GLM might be used. Modeling the number of third-party liability insurance claims with independent variables is the number of accidents using negative binomial distribution gives good results. This is indicated by the value of deviance approaching 1. Interpretation of the model indicates if the number of accidents increased 10%, then the number of claims will increase by 18.5%.     Keywords: insurance, third-party liability, clai