SAĞLIK SİGORTASI PRİM TAHMİNİNDE UÇ KIRPMA DUYARLILIĞI VE SHAP TABANLI AÇIKLANABİLİRLİK

Tuba Irmak

TR EN

SAĞLIK SİGORTASI PRİM TAHMİNİNDE UÇ KIRPMA DUYARLILIĞI VE SHAP TABANLI AÇIKLANABİLİRLİK

Öz

Sağlık sigortası primlerinin doğru tahmini, bireyler için gereksiz maliyetlerin önlenmesi ve sigorta şirketleri için sürdürülebilir kârlılığın sağlanması açısından kritik öneme sahiptir. Bu çalışmada, Kaggle'da açık erişimli "Insurance Premium Prediction" veri seti kullanılarak bireylerin yıllık sağlık sigortası prim tutarları makine öğrenmesi tabanlı regresyon yöntemleriyle tahmin edilmiştir. Doğrusal, örnek tabanlı, çekirdek tabanlı ve ağaç tabanlı topluluk yaklaşımlarından oluşan sekiz model eğitilmiş; performansları 5-katlı çapraz doğrulama ve ayrılmış test kümesi üzerinde karşılaştırılmıştır. En iyi sonuç 0,9007 R² değeriyle Gradient Boosting modelinden elde edilmiş, bu model için RandomizedSearchCV ile hiperparametre optimizasyonu yapılmış ve özellikle MAPE değerinde iyileşme sağlanmıştır. 'bmi' ve 'charges' değişkenlerine %1–%99 aralığında uç kırpma uygulanarak yürütülen duyarlılık analizinde, uç kırpmanın özellikle ağaç tabanlı modellerde R² değerlerini artırdığı ve hata metriklerini iyileştirdiği görülmüştür. SHAP analizi, prim tutarları üzerinde en etkili değişkenin sigara kullanım durumu olduğunu ortaya koymuştur. Bulguların genellenebilirliği bağımsız bir veri setinde sınanmış ve ağaç tabanlı topluluk modellerinin üstünlüğü doğrulanmıştır. Sonuçlar, uç değer işleme ile birleştirilen ağaç tabanlı topluluk modellerinin sigorta prim tahmini için güvenilir bir temel sunduğunu göstermektedir.

Anahtar Kelimeler

WINSORIZATION SENSITIVITY AND SHAP-BASED EXPLAINABILITY IN HEALTH INSURANCE PREMIUM PREDICTION

Öz

Accurate prediction of health insurance premiums is critical both for preventing unnecessary costs for individuals and for ensuring sustainable profitability for insurers. In this study, annual health insurance premiums are predicted using machine learning-based regression methods on the openly accessible "Insurance Premium Prediction" dataset shared on Kaggle. Eight models comprising linear, instance-based, kernel-based, and tree-based ensemble approaches were trained and compared through 5-fold cross-validation on the training set and on a held-out test set. The best performance was achieved by Gradient Boosting with an R² of 0.9007; hyperparameter optimization using RandomizedSearchCV yielded a further improvement, particularly in MAPE. A sensitivity analysis applying winsorization at the 1st–99th percentile range to the 'bmi' and 'charges' variables showed that outlier clipping increased R² values and improved error metrics, especially for tree-based models. SHAP analysis identified smoking status as the most influential variable on premium amounts. The generalizability of the findings was tested on an independent dataset, where tree-based ensemble models again outperformed the alternatives. These results suggest that tree-based ensembles combined with outlier treatment provide a reliable basis for insurance premium estimation.

Anahtar Kelimeler

Kaynakça

Ali, Z. A., Abduljabbar, Z. H., Tahir, H. A., Sallow, A. B., & Almufti, S. M. (2023). eXtreme gradient boosting algorithm with machine learning: A review. Academic Journal of Nawroz University, 12(2), 320-334. https://cir.nii.ac.jp/crid/1360021397008256384
Awad, M., & Khanna, R. (2015). Support vector regression. In Efficient learning machines: Theories, concepts, and applications for engineers and system designers (pp. 67-80). Berkeley, CA: Apress. https://doi.org/10.1007/978-1-4302-5990-9_4
Bau, Y. T., & Hanif, S. A. M. (2024). Comparative Analysis of Machine Learning Algorithms for Health Insurance Pricing. JOIV: International Journal on Informatics Visualization, 8(1), 481-491. https://doi.org/10.62527/joiv.8.1.2282
Bayo, A. K., Rafiu, A. B., Funmilayo, A. T., & Oluyemi, O. I. (2021). Investigating the impact of multicollinearity on linear regression estimates. Malaysian Journal of Computing (MJoC), 6(1), 698-714. https://doi.org/10.24191/mjoc.v6i1.10540
Bentéjac, C., Csörgő, A., & Martínez-Muñoz, G. (2021). A comparative analysis of gradient boosting algorithms. Artificial Intelligence Review, 54(3), 1937-1967. https://doi.org/10.1007/S10462-020-09896-5
Bhongade, A., Dubey, Y., Palsodkar, P., & Fulzele, P. (2024, December). Explainable AI Model for Medical Insurance Premium Prediction using Ensemble Learning with SHAP Analysis. In 2024 IEEE Pune Section International Conference (PuneCon) (pp. 1-6). IEEE. https://doi.org/10.1109/PuneCon63413.2024.10895118
Billa, M. M., & Nagpal, T. (2024). Medical insurance price prediction using machine learning. J. Electr. Syst, 20(7s), 2270-2279. https://doi.org/10.52783/jes.3962
Breiman, L. (2001). Random forests. Machine learning, 45(1), 5-32. https://doi.org/10.1023/a:1010933404324

Chen, T., & Guestrin, C. (2016, August). Xgboost: A scalable tree boosting system. In Proceedings of the 22nd acm sigkdd international conference on knowledge discovery and data mining (pp. 785-794). https://doi.org/10.1145/2939672.2939785
Dash, C. S. K., Behera, A. K., Dehuri, S., & Ghosh, A. (2023). An outliers detection and elimination framework in classification task of data mining. Decision Analytics Journal, 6, 100164. https://doi.org/10.1016/j.dajour.2023.100164
Dutta, S., Bose, P., & Bandyopadhyay, S. K. (2023). Forecasting health insurance premium using machine learning approaches. Asia-Pacific Journal of Science and Technology, 28(06), APST-28. https://doi.org/10.14456/apst.2023.96
Farhadi, Z., Belaghi, R. A., & Alma, O. G. (2019). Analysis of penalized regression methods in a simple linear model on the high-dimensional data. American Journal of Theoretical and Applied Statistics, 8(5), 185-192. https://doi.org/10.11648/j.ajtas.20190805.14
Franceschi, L., Donini, M., Perrone, V., Klein, A., Archambeau, C., Seeger, M., ... & Frasconi, P. (2025). Hyperparameter optimization in machine learning. Foundations and Trends® in Machine Learning, 18(6), 975-1109. https://doi.org/10.1561/2200000088
Friedman, J. H. (2001). Greedy function approximation: a gradient boosting machine. Annals of statistics, 29(5), 1189-1232. https://doi.org/10.1214/aos/1013203451
Friedman, J. H. (2002). Stochastic gradient boosting. Computational statistics & data analysis, 38(4), 367-378. https://doi.org/10.1016/S0167-9473(01)00065-2
Halder, R. K., Uddin, M. N., Uddin, M. A., Aryal, S., & Khraisat, A. (2024). Enhancing K-nearest neighbor algorithm: a comprehensive review and performance analysis of modifications. Journal of Big Data, 11(1), 113. https://doi.org/10.1186/s40537-024-00973-y
Hu,Y. (2024). Comparison and Analysis of the Effectiveness of Linear Regression, Decision Tree, and Random Forest Models for Health Insurance Premium Forecasting. Advances in Economics, Management and Political Sciences, 79, 347-353. https://doi.org/10.54254/2754-1169/79/20241754
Insurance Premium Prediction Dataset, Erişim Tarihi: 11.09.2025. Erişim Bağlantısı: https://www.kaggle.com/datasets/noordeen/insurance-premium-prediction
Irmak, T. (2025, July). Analysis of Order Cancellation Rates and Feature Weighting with Machine Learning: An E-Commerce Case Study. In International Conference on Intelligent and Fuzzy Systems (pp. 763-770). Cham: Springer Nature Switzerland. https://doi.org/10.1007/978-3-031-98304-7_82
Kapse, M., Sharma, V., Vidhale, R., & Vellanki, V. (2025). Customization of health insurance premiums using machine learning and explainable AI. International Journal of Information Management Data Insights, 5(1), 100328. https://doi.org/10.1016/j.jjimei.2025.100328
Kaushik, K., Bhardwaj, A., Dwivedi, A. D., & Singh, R. (2022). Machine learning-based regression framework to predict health insurance premiums. International journal of environmental research and public health, 19(13), 7898. https://doi.org/10.3390/ijerph19137898
Khajeh, M., Ghaffari-Moghaddam, M., Piri, J., Barkhordar, A., & Şenol, H. (2026). Sustainable Beetroot Bioactives Extraction with NADES-MAE: A Spider Algorithm-Optimized Hybrid SVR-XGBoost Model. Applied Food Research, 101940. https://doi.org/10.1016/j.afres.2026.101940
Kulkarni, M., Mesharam, D. D., Patil, B., More, R., Sharma, M., & Patange, P. (2022). Medical insurance cost prediction using machine learning. International Journal for Research in Applied Science & Engineering Technology, 10. https://doi.org/10.22214/ijraset.2022.47923
Liaw, A., & Wiener, M. (2002). Classification and regression by randomForest. R news, 2(3), 18-22.
Lin, K., Diab, L. S., Zhang, Y., & ELshreif, A. S. (2026). Analyzing the performance of machine learning techniques: Applications in dance education and the radiation sector. Journal of Radiation Research and Applied Sciences, 19(3), 102522. https://doi.org/10.1016/j.jrras.2026.102522
Liu, Y., & Zhang, S. (2026). Application of Machine Learning Algorithms for Medical Insurance Cost Prediction and Feature Analysis. Emerging Markets Finance and Trade, 62(1), 1-17. https://doi.org/10.1080/1540496X.2025.2522233
Lundberg, S. M., & Lee, S. I. (2017). A unified approach to interpreting model predictions. Advances in neural information processing systems, 30. https://doi.org/10.48550/arXiv.1705.07874
Mahesh, M., Priyatham, B. S., SaiKumar, G., & Reddy, K. V. V. (2025, May). Predictive Analytics for Medical Insurance Premiums. In 2025 5th International Conference on Pervasive Computing and Social Networking (ICPCSN) (pp. 303-309). IEEE. https://doi.org/10.1109/ICPCSN65854.2025.11034884
Natekin, A., & Knoll, A. (2013). Gradient boosting machines, a tutorial. Frontiers in neurorobotics, 7, 21. https://doi.org/10.3389/fnbot.2013.00021
Orji, U., & Ukwandu, E. (2024). Machine learning for an explainable cost prediction of medical insurance. Machine learning with applications, 15, 100516. https://doi.org/10.1016/j.mlwa.2023.100516
Ponce‐Bobadilla, A. V., Schmitt, V., Maier, C. S., Mensing, S., & Stodtmann, S. (2024). Practical guide to SHAP analysis: Explaining supervised machine learning model predictions in drug development. Clinical and translational science, 17(11), e70056. https://doi.org/10.1111/cts.70056
Rao, D. M., Somalaraju, L. K. R., Bodagala, V. S. M., & Edapalapati, V. B. J. S. (2025, May). Predicting Medical Insurance Costs: A Machine Learning Approach for Smarter Risk Assessment. In 2025 International Conference on Computational Robotics, Testing and Engineering Evaluation (ICCRTEE) (pp. 1-7). IEEE. https://doi.org/10.1109/ICCRTEE64519.2025.11052963
Saputro, D. R. S., Wahyu, N. L., & Widyaningsih, Y. (2025). Performance of Ridge Regression, Least Absolute Shrinkage and Selection Operator, and Elastic Net in Overcoming Multicollinearity. Journal of Multidisciplinary Applied Natural Science, 5(2), 370-382. https://doi.org/10.47352/jmans.2774-3047.251
Saraiva, F. M. G., Guimarães, A. C. R., da Silva Júnior, O. S., Monteiro, S. N., & Coelho, L. M. (2026). Performance analysis of Gradient Boosting Regressor Using K-fold and Group K-fold for Predicting Permanent Deformation in Tropical Soils. Journal of Materials Research and Technology. https://doi.org/10.1016/j.jmrt.2025.12.152
Sharma, A., Malik, A., Sharma, S., Batra, I., & Hussain, M. (2024, November). Empowering Healthcare Accessibility: Machine Learning-Driven Predictive Modeling for Health Insurance Premiums in India. In 2024 IEEE 9th International Conference on Engineering Technologies and Applied Sciences (ICETAS) (pp. 1-7). IEEE. https://doi.org/10.1109/ICETAS62372.2024.11119999
Sijie, L., Sia, F., Alfred, R., & Moung, E. G. (2024, August). Ensemble Machine Learning Method for Health Insurance Premium Prediction. In 2024 IEEE International Conference on Artificial Intelligence in Engineering and Technology (IICAIET) (pp. 642-646). IEEE. https://doi.org/10.1109/IICAIET62352.2024.10730721
Smola, A. J., & Schölkopf, B. (2004). A tutorial on support vector regression. Statistics and computing, 14(3), 199-222. https://doi.org/10.1023/B:STCO.0000035301.49549.88
Topal, B. (2024). Özel Sağlık Sigortası Primleri Üzerinde Makroekonomik Göstergelerin Etkisi. Yüksek Lisans Tezi, Sağlık Yönetimi Anabilim Dalı, Düzce Üniversitesi.
Usman, M., SIS, D., & Alhaji, B. B. (2021). Comparing the Prediction Accuracy of Ridge, Lasso and Elastic Net Regression Models with Linear Regression using Breast Cancer Data. Bayero Journal of Pure & Applied Sciences, 14(2), 134-149. https://doi.org/10.4314/bajopas.v14i2.16
Zhang, Z. (2016). Introduction to machine learning: k-nearest neighbors. Annals of translational medicine, 4(11), 218. https://doi.org/10.21037/atm.2016.03.37
Zubedi, F., Sartono, B., & Notodiputro, K. A. (2022). Implementation of Winsorizing and random oversampling on data containing outliers and unbalanced data with the random forest classification method. Jurnal Natural, 22(2), 108-116. https://doi.org/10.24815/jn.v22i2.25499

Ayrıntılar

Birincil Dil

Türkçe

Konular

Makine Öğrenme (Diğer)

Bölüm

Araştırma Makalesi

Yazarlar

Tuba Irmak ^*
0000-0003-4749-5226
Türkiye

Yayımlanma Tarihi

3 Eylül 2026

Gönderilme Tarihi

9 Şubat 2026

Kabul Tarihi

28 Temmuz 2026

Yayımlandığı Sayı

Yıl 2026 Cilt: 29 Sayı: 3

IZ

https://izlik.org/JA45KL84FW

Kaynak Göster

RIS / Bibtex

APA

Irmak, T. (2026). SAĞLIK SİGORTASI PRİM TAHMİNİNDE UÇ KIRPMA DUYARLILIĞI VE SHAP TABANLI AÇIKLANABİLİRLİK. Kahramanmaraş Sütçü İmam Üniversitesi Mühendislik Bilimleri Dergisi, 29(3), 1042-1061. https://izlik.org/JA45KL84FW

AMA

1.Irmak T. SAĞLIK SİGORTASI PRİM TAHMİNİNDE UÇ KIRPMA DUYARLILIĞI VE SHAP TABANLI AÇIKLANABİLİRLİK. Kahramanmaraş Sütçü İmam Üniversitesi Mühendislik Bilimleri Dergisi. 2026;29(3):1042-1061. https://izlik.org/JA45KL84FW

Chicago

Irmak, Tuba. 2026. “SAĞLIK SİGORTASI PRİM TAHMİNİNDE UÇ KIRPMA DUYARLILIĞI VE SHAP TABANLI AÇIKLANABİLİRLİK”. Kahramanmaraş Sütçü İmam Üniversitesi Mühendislik Bilimleri Dergisi 29 (3): 1042-61. https://izlik.org/JA45KL84FW.

EndNote

Irmak T (01 Eylül 2026) SAĞLIK SİGORTASI PRİM TAHMİNİNDE UÇ KIRPMA DUYARLILIĞI VE SHAP TABANLI AÇIKLANABİLİRLİK. Kahramanmaraş Sütçü İmam Üniversitesi Mühendislik Bilimleri Dergisi 29 3 1042–1061.

IEEE

[1]T. Irmak, “SAĞLIK SİGORTASI PRİM TAHMİNİNDE UÇ KIRPMA DUYARLILIĞI VE SHAP TABANLI AÇIKLANABİLİRLİK”, Kahramanmaraş Sütçü İmam Üniversitesi Mühendislik Bilimleri Dergisi, c. 29, sy 3, ss. 1042–1061, Eyl. 2026, [çevrimiçi]. Erişim adresi: https://izlik.org/JA45KL84FW

ISNAD

Irmak, Tuba. “SAĞLIK SİGORTASI PRİM TAHMİNİNDE UÇ KIRPMA DUYARLILIĞI VE SHAP TABANLI AÇIKLANABİLİRLİK”. Kahramanmaraş Sütçü İmam Üniversitesi Mühendislik Bilimleri Dergisi 29/3 (01 Eylül 2026): 1042-1061. https://izlik.org/JA45KL84FW.

JAMA

1.Irmak T. SAĞLIK SİGORTASI PRİM TAHMİNİNDE UÇ KIRPMA DUYARLILIĞI VE SHAP TABANLI AÇIKLANABİLİRLİK. Kahramanmaraş Sütçü İmam Üniversitesi Mühendislik Bilimleri Dergisi. 2026;29:1042–1061.

MLA

Irmak, Tuba. “SAĞLIK SİGORTASI PRİM TAHMİNİNDE UÇ KIRPMA DUYARLILIĞI VE SHAP TABANLI AÇIKLANABİLİRLİK”. Kahramanmaraş Sütçü İmam Üniversitesi Mühendislik Bilimleri Dergisi, c. 29, sy 3, Eylül 2026, ss. 1042-61, https://izlik.org/JA45KL84FW.

Vancouver

1.Tuba Irmak. SAĞLIK SİGORTASI PRİM TAHMİNİNDE UÇ KIRPMA DUYARLILIĞI VE SHAP TABANLI AÇIKLANABİLİRLİK. Kahramanmaraş Sütçü İmam Üniversitesi Mühendislik Bilimleri Dergisi [Internet]. 01 Eylül 2026;29(3):1042-61. Erişim adresi: https://izlik.org/JA45KL84FW