Perbandingan Metode IQR vs Isolation Forest dalam Deteksi Outlier pada Data Klaim Asuransi di PT Askrindo Cabang Semarang 358 372
DOI:
https://doi.org/10.26623/transformatika.v23i2.13726Kata Kunci:
Deteksi Anomali, Iqr, Hutan Terpencil, Klaim Asuransi, Penyandang data anomaliAbstrak
Penelitian ini bertujuan untuk membandingkan dua pendekatan yang umum digunakan dalam deteksi outlier, yaitu metode statistik Interquartile Range (IQR) dan metode machine learning Isolation Forest, studi kasus data klaim asuransi di PT Askrindo Cabang Semarang. Data yang dianalisis mencakup nilai klaim dan durasi penyelesaian. Metode IQR secara univariat menemukan 9 klaim yang terklasifikasi sebagai outlier, yang terdiri atas 5 outlier nilai klaim, 3 outlier durasi penyelesaian, serta 1 outlier ganda. Sedangkan Isolation Forest secara multivariat dengan contamination 0,05 mendeteksi 6 klaim anomali, termasuk durasi penyelesaian sangat lama dan pola kombinasi variabel yang tidak wajar. Hasilnya IQR efektif untuk deteksi cepat nilai ekstrem, sedangkan Isolation Forest lebih cocok dalam mengenali anomali kompleks antar variabel. Oleh karena itu, penggunaan metode IQR maupun Isolation Forest dapat disesuaikan dengan tujuan analisis karakteristik data.
Unduhan
Referensi
[1] D. L. Whaley, The Interquartile Range: Theory and Estimation, Electron. Theses Diss., p. 54, 2005, [Online]. Available: http://dc.etsu.edu/etd%0Ahttp://dc.etsu.edu/etd
[2] D. Herurkar, S. Palacio, A. Anwar, J. Hees, and A. Dengel, Fin-Fed-OD: Federated Outlier Detection on Financial Tabular Data, 2024, [Online]. Available: http://arxiv.org/abs/2404.14933
[3] S. Buschjäger, P. J. Honysz, and K. Morik, Randomized outlier detection with trees, Int. J. Data Sci. Anal., vol. 13, no. 2, pp. 91–104, 2022, doi: 10.1007/s41060-020-00238-w.
[4] A. Capozzi et al., FlowSeries: Anomaly Detection in Financial Transaction Flows, Stud. Comput. Intell., vol. 1189 SCI, pp. 29–40, 2025, doi: 10.1007/978-3-031-82435-7_3.
[5] P. K. Dongre, V. Patel, U. Bhoi, and N. N. Maltare, An outlier detection framework for Air Quality Index prediction using linear and ensemble models, Decis. Anal. J., vol. 14, no. July 2024, p. 100546, 2025, doi: 10.1016/j.dajour.2025.100546.
[6] Dr. Velmurugan. K, Mr. K. Pazhanivel, Divyasree. R, Gowtham. E, and Guruharan. S, Data Driven Analysis of Insurance Claims Using Machine Learning Algorithm, Int. J. Adv. Res. Sci. Commun. Technol., pp. 269–273, 2023, doi: 10.48175/ijarsct-9689.
[7] I. Forest and I. Forest, Isolation Forest算法实现详解 （ 点击阅读原文有惊喜哦 ）, vol. 9818, no. 1, pp. 1–15, 2021.
[8] R. Items, W. Rose, W. Rose, T. If, and W. Rose, Encyclopedia of business analytics and optimization, Choice Rev. Online, vol. 52, no. 02, pp. 52-0606-52–0606, 2014, doi: 10.5860/choice.52-0606.
[9] P. O. C. Sari and S. Suharjito, Outlier Detection in Inpatient Claims Using DBSCAN and K-Means, J. Tek. Inform., vol. 15, no. 1, pp. 1–10, 2022, doi: 10.15408/jti.v15i1.25682.
[10] B. Avanzi, M. Lavender, G. Taylor, and B. Wong, Detection and treatment of outliers for multivariate robust loss reserving, Ann. Actuar. Sci., vol. 18, no. 1, pp. 102–125, 2024, doi: 10.1017/S1748499523000155.
[11] M. Ok, S. Klüttermann, and E. Müller, Exploring the Impact of Outlier Variability on Anomaly Detection Evaluation Metrics, 2024, [Online]. Available: http://arxiv.org/abs/2409.15986
[12] R. Selina, Y. Yanti, and A. Suryono, Analisis Yuridis Tanggung Jawab Perusahaan Asuransi Terhadap Asuransi Jiwa Kredit Rahma Selina Yustika Yanti Arief Suryono K / Pdt / 2019 ). Penolakan pembayaran klaim meninggal dunia dilakukan oleh pihak PT . Asuransi Jiwa Inhealth Indonesia kepada pemeg, no. 1, 2024.
[13] M. Sri Lakshmi, G. Rajavikram, V. Dattatreya, B. Swarna Jyothi, S. Patil, and M. Bhavsingh, Evaluating the Isolation Forest Method for Anomaly Detection in Software-Defined Networking Security, J. Electr. Syst., vol. 19, no. 4, pp. 279–297, 2023, doi: 10.52783/jes.639.
[14] C. Shao, X. Du, J. Yu, and J. Chen, Cluster-Based Improved Isolation Forest, Entropy, vol. 24, no. 5, 2022, doi: 10.3390/e24050611.
Unduhan
Diterbitkan
Terbitan
Bagian
Lisensi
Hak Cipta (c) 2026 jovita wayan ramadhani, susanto
Artikel ini berlisensi Creative Commons Attribution 4.0 International License.
Authors who publish with this journal agree to the following terms:
- Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work.
Transformatika is licensed under a Creative Commons Attribution 4.0 International License.
