Pemodelan Status Gagal Bayar pada Asuransi Kebakaran PT BRI Asuransi Indonesia KCU Surabaya Menggunakan Regresi Logistik Biner

Setyadi, Angelica Benedict and Nuzula, Firdausi (2024) Pemodelan Status Gagal Bayar pada Asuransi Kebakaran PT BRI Asuransi Indonesia KCU Surabaya Menggunakan Regresi Logistik Biner. Project Report. [s.n.], [s.l.]. (Unpublished)

Text 5006211021_5006211053-Project_Report.pdf - Accepted Version Restricted to Repository staff only Download (7MB) | Request a copy

Abstract

Kerja Praktik merupakan salah satu mata kuliah wajib yang bertujuan memberikan pengalaman nyata dalam penerapan ilmu aktuaria di dunia industri. Kegiatan ini dilaksanakan di PT BRI Asuransi Indonesia Cabang Surabaya selama satu bulan dengan fokus pada analisis status gagal bayar asuransi kebakaran menggunakan metode regresi logistik biner. Data yang dianalisis mencakup 475 data outstanding premium yang memuat informasi status pembayaran, relationship marketing, periode polis, nilai premi, dan nilai pertanggungan. Analisis dilakukan melalui statistika deskriptif, pemeriksaan multikolinearitas, pembentukan model regresi logistik biner, uji signifikansi parameter, uji kesesuaian model, dan evaluasi ketepatan klasifikasi. Hasil penelitian menunjukkan bahwa variabel yang berpengaruh signifikan terhadap status gagal bayar adalah relationship marketing, periode polis, dan nilai premi, sedangkan nilai pertanggungan tidak signifikan. Model yang dihasilkan memiliki ketepatan klasifikasi sebesar 77,9%, sehingga dapat dimanfaatkan perusahaan untuk mengidentifikasi potensi risiko gagal bayar dan menyusun strategi mitigasi risiko yang lebih efektif. Penelitian ini diharapkan memberikan kontribusi bagi peningkatan pengelolaan risiko keuangan perusahaan.
===================================================================================================================================
The Internship Program is a compulsory course aimed at providing hands-on experience in applying actuarial science in the insurance industry. This internship was carried out at PT BRI Asuransi Indonesia, Surabaya Branch, for one month, focusing on analyzing the default status of fire insurance policies using binary logistic regression. The dataset consisted of 475 outstanding premium records containing payment status, relationship marketing, policy period, premium value, and sum insured. The analysis involved descriptive statistics, multicollinearity tests, binary logistic regression modeling, parameter significance testing, goodness-of-fit assessment, and classification accuracy evaluation. The results indicate that relationship marketing, policy period, and premium value significantly affect the likelihood of default, while the sum insured does not have a significant effect. The developed model achieved a classification accuracy of 77.9%, enabling the company to better identify potential default risks and implement more effective risk mitigation strategies. This study is expected to contribute to improving the company’s financial risk management.

Item Type:	Monograph (Project Report)
Uncontrolled Keywords:	kerja praktik, asuransi kebakaran, gagal bayar, regresi logistik biner, ketepatan klasifikasi, internship, fire insurance, default, binary logistic regression, classification accuracy.
Subjects:	Q Science > QA Mathematics > QA276 Mathematical statistics. Time-series analysis. Failure time data analysis. Survival analysis (Biometry) Q Science > QA Mathematics > QA278.2 Regression Analysis. Logistic regression
Divisions:	Faculty of Mathematics, Computation, and Data Science > Actuaria > 94203-(S1) Undergraduate Thesis
Depositing User:	Firdausi Nuzula
Date Deposited:	31 Jul 2025 02:21
Last Modified:	31 Jul 2025 02:21
URI:	http://repository.its.ac.id/id/eprint/123246

Actions (login required)

View Item