Membangun pemodelan kerugian dengan menggunakan generalized linear mixed models untuk asuransi kendaraan bermotor = Modeling vehicle insurance losses using generalized linear mixed models

Budhyanto, Johana Daniella (2024) Membangun pemodelan kerugian dengan menggunakan generalized linear mixed models untuk asuransi kendaraan bermotor = Modeling vehicle insurance losses using generalized linear mixed models. Bachelor thesis, Universitas Pelita Harapan.

[img]
Preview
Text (Title)
Title.pdf
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (247kB) | Preview
[img]
Preview
Text (Abstract)
Abstract.pdf
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (1MB) | Preview
[img]
Preview
Text (ToC)
ToC.pdf
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (590kB) | Preview
[img]
Preview
Text (Chapter1)
Chapter1.pdf
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (568kB) | Preview
[img] Text (Chapter2)
Chapter2.pdf
Restricted to Registered users only
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (678kB)
[img] Text (Chapter3)
Chapter3.pdf
Restricted to Registered users only
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (710kB)
[img] Text (Chapter4)
Chapter4.pdf
Restricted to Registered users only
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (979kB)
[img] Text (Chapter5)
Chapter5.pdf
Restricted to Registered users only
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (562kB)
[img]
Preview
Text (Bibliography)
Bibliography.pdf
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (564kB) | Preview
[img] Text (Appendices)
Appendices.pdf
Restricted to Repository staff only
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (7MB)

Abstract

Sebuah perusahaan asuransi perlu menentukan harga kontrak atau premi secara memadai sehingga diperlukan model yang dapat memprediksi besar kerugian yang mungkin terjadi di masa depan. Penelitian ini bertujuan untuk memproyeksikan besar kerugian yang mungkin terjadi dan juga membandingkan performa dari Generalized Linear Models (GLM) dan Generalized Linear Mixed Models (GLMM). Terdapat tiga jenis data yang akan digunakan untuk membuat model, yakni seluruh data, data besar klaim 0 - 500, dan data besar klaim positif. Dari masing-masing jenis data akan dilakukan pemodelan frekuensi dan besar klaim dengan GLM dan GLMM. Untuk seluruh data, melalui analisis MSE dan MAE diperoleh bahwa kombinasi GLM Binomial dengan GLM Gamma memberikan hasil yang terbaik. Untuk data klaim 0 - 500, melalui analisis MSE diperoleh bahwa kombinasi GLM Binomial dengan GLM Gamma memberikan hasil terbaik. Melalui analisis MAE diperoleh bahwa kombinasi GLM Poisson dengan GLM Inverse Gaussian memberikan hasil terbaik. Untuk data klaim positif, melalui analisis MSE dan MAE diperoleh bahwa kombinasi GLM Negatif Binomial dengan GLM Gamma memberikan hasil terbaik. Selain itu, melalui analisis faktor risiko diperoleh bahwa pengendara laki-laki, kelompok pengendara berusia paling muda, kelompok kendaraan berusia tua, kelompok kendaraan dengan nilai yang tinggi, serta jenis kendaraan Coupe memberikan ekspektasi kerugian yang lebih besar daripada variabel yang lainnya. / An insurance company needs to determine contract or premium prices adequately, so it requires models that can predict potential future losses. This research aims to project potential losses and compare the performance of Generalized Linear Models (GLM) and Generalized Linear Mixed Models (GLMM). There are three types of data that will be used to create the model consisting of entire dataset, data on claims between 0 and 500 dollars, and data for positive claims. Frequency and severity of claims will be modeled for each data type using both GLM and GLMM. For the entire dataset, an analysis of Mean Squared Error (MSE) and Mean Absolute Error (MAE) reveals that the combination of Binomial GLM with Gamma GLM performs the best. For the 0-500 claims data, an MSE analysis shows that the combination of Binomial GLM with Gamma GLM performs the best. An MAE analysis indicates that the combination of Poisson GLM with Inverse Gaussian GLM performs the best. In the case of positive claims data, an analysis of MSE and MAE reveals that the combination of Negative Binomial GLM with Gamma GLM performs the best. Furthermore, a risk factor analysis shows that male drivers, the youngest age group of drivers, older vehicles, high values vehicles, and Coupe-type vehicles contribute to higher expected losses compared to other variables.

Item Type: Thesis (Bachelor)
Creators:
CreatorsNIMEmail
Budhyanto, Johana DaniellaNIM01112200018johanadaniellabudhyanto@gmail.com
Contributors:
ContributionContributorsNIDN/NIDKEmail
Thesis advisorSaputra, Kie Van IvankyNIDN0401038203kie.saputra@uph.edu
Uncontrolled Keywords: generalized linear mixed models; generalized linear models; asuransi kendaraan bermotor
Subjects: Q Science > QA Mathematics
Divisions: University Subject > Current > Faculty/School - UPH Karawaci > Faculty of Science and Technology > Mathematics
Current > Faculty/School - UPH Karawaci > Faculty of Science and Technology > Mathematics
Depositing User: Johana Daniella Budhyanto
Date Deposited: 31 Jan 2024 16:42
Last Modified: 31 Jan 2024 16:42
URI: http://repository.uph.edu/id/eprint/61303

Actions (login required)

View Item View Item