SỐ 241 | Tháng 4/2026
Mô hình đánh giá tín dụng: Tiếp cận từ tổn thất phân loại sai
Bùi Thị Thiện Mỹ
Tóm tắt:
Mục tiêu nghiên cứu: Trên bộ dữ liệu gồm 9.524 khách hàng có quan hệ tín dụng với một ngân hàng, bài báo đề xuất một mô hình đánh giá tín dụng (ĐGTD) nhị phân có tính đến tổn thất khi phân loại sai. Mô hình được kỳ vọng đảm bảo hiệu quả phân loại và tối thiểu hoá tổn thất mà ngân hàng gánh chịu.
Thiết kế nghiên cứu/phương pháp/tiếp cận: Bài viết thiết kế một mô hình hồi quy Logistic có tính đến tổn thất khi phân loại sai. Trong đó, các giá trị tổn thất khi phân loại sai phụ thuộc vào từng khách hàng.
Kết quả nghiên cứu chính: Khi so sánh với các mô hình thông dụng, mô hình đề xuất có hiệu quả phân loại tốt trên cả nhóm khách hàng tốt và xấu đồng thời có mức cải thiện tổn thất là tốt nhất.
Giá trị đóng góp: Tiếp cận từ tổn thất khi phân loại sai là một hướng tiếp cận hợp lý với bài toán ĐGTD. Mô hình đề xuất có hiệu quả phân loại tốt, dễ sử dụng và không đòi hỏi nhiều các hiệu chỉnh lên mô hình hồi quy Logistic truyền thống.
Tài liệu tham khảo:
- Bahnsen, A. C., Aouada, D., & Ottersten, B. (2014). Example-dependent cost-sensitive logistic regression for credit scoring. Proceedings - 2014 13th International Conference on Machine Learning and Applications, ICMLA 2014, 263-269. https://doi.org/10.1109/ICMLA.2014.48.
- Bahnsen, A. C., Aouada, D., & Ottersten, B. (2015). Example-dependent cost-sensitive decision trees. Expert Systems with Applications, 42(19), 6609-6619. https://doi.org/10.1016/j.eswa.2015.04.042.
- Batista, G. E., Prati, R. C., & Monard, M. C. (2004). A study of the behavior of several methods for balancing machine learning training data. ACM SIGKDD Explorations Newsletter, 6(1), 20-29.
- Brown, I., & Mues, C. (2012). An experimental comparison of classification algorithms for imbalanced credit scoring data sets. Expert Systems with Applications, 39(3), 3446-3453.
- Castro, C. L., & Braga, A. P. (2013). Novel cost-sensitive approach to improve the multilayer perceptron performance on imbalanced data. IEEE Transactions on Neural Networks and Learning Systems, 24(6), 888-899.
- Chawla, N. V., Bowyer, K. W., Hall, L. O., & Kegelmeyer, W. P. (2002). SMOTE: Synthetic minority over-sampling technique. Journal of Artificial Intelligence Research, 16, 321-357. https://doi.org/10.1613/jair.953.
- Datta, S., & Das, S. (2015). Near-Bayesian support vector machines for imbalanced data classification with equal or unequal misclassification costs. Neural Networks, 70, 39-52.
- Drummond, C., & Holte, R. C. (2003). C4.5, class imbalance, and cost sensitivity: Why under-sampling beats over-sampling. In Workshop on Learning from Imbalanced Datasets II (pp. 1-8), held in conjunction with ICML 2003.
- D’Addabbo, A., & Maglietta, R. (2015). Parallel selective sampling method for imbalanced and large data classification. Pattern Recognition Letters, 62, 61-67.
- Elkan, C. (2001). The foundations of cost-sensitive learning. International Joint Conference on Artificial Intelligence, 17(1), 973-978.
- Fan, W., Stolfo, S. J., Zhang, J., & Chan, P. K. (1999). AdaCost: Misclassification cost-sensitive boosting. In Proceedings of the Sixteenth International Conference on Machine Learning (pp. 97-105). Morgan Kaufmann.
- Fernández, A., García, S., Galar, M., Prati, R. C., Krawczyk, B., & Herrera, F. (2018). Learning from imbalanced data sets. Springer. https://doi.org/10.1007/978-3-319-98074-4.
- Gareth, J., Daniela, W., Trevor, H., & Robert, T. (2013). An introduction to statistical learning: with applications in R. Spinger.
- Haixiang, G., Yijing, L., Shang, J., Mingyun, G., Yuanyue, H., & Bing, G. (2017). Learning from class-imbalanced data: Review of methods and applications. Expert Systems with Applications, 73, 220-239. https://doi.org/10.1016/j.eswa.2016.12.035.
- Iranmehr, A., Masnadi-Shirazi, H., & Vasconcelos, N. (2019). Cost-sensitive support vector machines. Neurocomputing, 343, 50-64.
- Jabeur, S. Ben, Sadaaoui, A., Sghaier, A., & Aloui, R. (2020). Machine learning models and cost-sensitive decision trees for bond rating prediction. Journal of the Operational Research Society, 71(8), 1161-1179. https://doi.org/10.1080/01605682.2019.1581405.
- Kuznetsov, S. O., Masyutin, A., & Ageev, A. (2020). Are you a good borrower? Mining interpretable pattern structures in credit scoring. Asian Journal of Economics and Banking, 4(3), 67–85. https://doi.org/10.1108/ajeb-08-2020-0056
- López, V., Del Río, S., Benítez, J. M., & Herrera, F. (2015). Cost-sensitive linguistic fuzzy rule based classification systems under the MapReduce framework for imbalanced big data. Fuzzy Sets and Systems, 258, 5-38.
- Loyola-González, O., Martínez-Trinidad, J. F., Carrasco-Ochoa, J. A., & García-Borroto, M. (2016). Study of the impact of resampling methods for contrast pattern based classifiers in imbalanced databases. Neurocomputing, 175, 935-947. https://doi.org/10.1016/j.neucom.2015.04.120.
- Ma, Y., Zhao, K., Wang, Q., & Tian, Y. (2020). Incremental cost-sensitive support vector machine with linear-exponential loss. IEEE Access, 8, 149899-149914.
- Marqués, A. I., Garcia, V., & Sánchez, J. S. (2013). On the suitability of resampling techniques for the class imbalance problem in credit scoring. Journal of the Operational Research Society, 64(7), 1060-1070.
- Moepya, S. O., Akhoury, S. S., & Nelwamondo, F. V. (2014). Applying cost-sensitive classification for financial fraud detection under high class-imbalance. 2014 IEEE International Conference on Data Mining Workshop (ICDMW), 183-192. https://doi.org/10.1109/ICDMW.2014.141.
- Onay, C., & Öztürk, E. (2018). A review of credit scoring research in the age of Big Data. Journal of Financial Regulation and Compliance, 26(3), 382-405. https://doi.org/10.1108/JFRC-06-2017-0054.
- Petrides, G., Moldovan, D., Coenen, L., Guns, T., & Verbeke, W. (2022). Cost-sensitive learning for profit-driven credit scoring. Journal of the Operational Research Society, 73(2), 338-350. https://doi.org/10.1080/01605682.2020.1843975.
- Reddy, K. S., Yadav, R. S., & Agarwal, A. (2025). Credit exposures and systemic risk in Indian banks. Asian Journal of Economics and Banking, 9(2), 222–239. https://doi.org/10.1108/ajeb-12-2024-0148
- Shen, F., Wang, R., & Shen, Y. (2020). A cost-sensitive logistic regression credit scoring model based on multi-objective optimization approach. Technological and Economic Development of Economy, 26(2), 405-429. https://doi.org/10.3846/tede.2019.11337.
- Sheng, V. S., & Ling, C. X. (2006). Thresholding for making classifiers cost-sensitive. Proceedings of the National Conference on Artificial Intelligence, 1, 476-481.
- Sushma, S. J., Kumar, S. C. P., & Assegie, T. A. (2023). A cost-sensitive logistic regression model for breast cancer detection. The Imaging Science Journal, 70(1), 10-18.
A Credit Scoring Model: Cost-Sensitive Learning Approach
Abstract:
Purpose: Based on a dataset of 9524 customers with credit relationships at a bank, the paper proposes a binary credit scoring model that incorporates misclassification cost. The model is expected to ensure classification performance while minimizing the bank's expected costs.
Design/methodology/approach: The study employs a Logistic regression model that accounts for misclassification cost. In this framework, the costs depend on each customer.
Findings: Compared with common models, the proposed model demonstrates good classification performance for both default and non-default customers, while achieving the best improvement in loss reduction.
Originality/value: Approaching cost-sensitive learning to credit scoring is a reasonable solution. The proposed model offers effective classification, ease of use, and requires minimal adjustments to the traditional Logistic regression framework.