No image available for this title

Skripsi

Claim severity modeling using the generalized linear model in motor vehicle insurance / Ferdi Maulana Ikhsan

Abstrak
Claim frequency and severity modeling are the key aspects that insurance companies can do to ensure the premium tariff is appropriate for the insurance user. A widely used method for modeling these two data is the generalized linear model. Compared to the linear regression model the Generalized Linear Model offers the ability to choose the distribution data other than the normal distribution. Additionally a link function in the Generalized Linear Model enables a connection between the regression model and the expected value of the response variable. However the performance of a Generalized Linear Model heavily depends on the choice of distribution and the link function. This study aims to identify the best performance model from the distribution and link functions evaluated based on the Akaike Information Criteria (AIC) value. The distributions used are the Gamma distribution and Inverse Gaussian distribution. Each model will use the canonical link function of each distribution and log link function. The procedure to achieve those goals includes data cleaning multicollinearity checking creating four models based on the criteria before and evaluating the performances of each model. The backward stepwise is applied in each model to ensure that the model includes only the most significant variables. The result shows that the Generalized Linear Model Inverse Gaussian that uses the log link function is the best model with 7760.8 AIC value. This research has concluded that the Inverse Gaussian distribution has better performance to implement than the Gamma distribution for modeling auto insurance claim severity under both the canonical and log link functions.

Informasi Detail
DDC
SKRIPSI DIGITAL
Prodi
Universitas Negeri Malang. Program Studi Matematika, .
Deskripsi Fisik
xii, 58 hlm. : ilus.
Bahasa
No Reg
3050/RS/25
Edisi
Skripsi (Sarjana)--Universitas Negeri Malang. 2025
Subjek
I. KEPARAHAN KLAIM - MODEL LINIER UMUM
II. FUNGSI TAUTAN
III. CLAIM SEVERITY - GENERAL LINEAR MODEL
Pembimbing
1. Lita Wulandari Aeli, M.si
Lampiran Berkas