DETERMINATION OF PURE PREMIUM USING GENERALIZED LINEAR MODEL (GLM) CASE STUDY: INPATIENT HEALTH INSURANCE

• المؤلف الرئيسي: MARYAM (NIM : 10109105); Pembimbing : Dumaria. R. Tampubolon. Ph.D, ISA
• التنسيق: Final Project
• اللغة: Indonesia
• الوصول للمادة أونلاين: https://digilib.itb.ac.id/gdl/view/16113

In this final project (skripsi), the pure premium for inpatient health insurance of an insurance company, is determined. The determination of the pure premium is dependent on claim size (severity) and claim numbers (claim frequency). The severity and frequency of claims are modeled using Generalized Linear Model (GLM) with log link function. Using the obtained models, the pure premium is determined using a compound model. The data used as a case study is inpatient health insurance data for underwriting year 2012 of an insurance company. Using the Kolmogorov-Smirnov test, for a significance level of (alpha) 5% the response variable, claims paid by the insurance company, does not follow any distributions belonging to the exponential family distributions. For the purpose of the objectives of the research, the claims severity are assumed to follow Inverse Gaussian and Gamma distributions. It is found that, at (alpha) 5% significance level, claims severity is determined by “length of stay”, “ICD codes” and “types of classes/services at the hospital”. For the claims frequency, at (alpha) 5%, significance level, the chi-square test showed that the claims frequency follows a negative binomial distribution. The model shows that the claims frequency is determined by “age” and “marital status” of the patient.