Veri madenciliği yöntemleriyle kredi skor tahmini: Performans karşılaştırması ve analizi

Öz

Kredi skoru tahmini, finansal kuruluşların kredi riskini etkin şekilde yönetmeleri ve sürdürülebilir kârlılık sağlamaları açısından kritik bir öneme sahiptir. Sağlıklı kredi kararlarının alınabilmesi için geçmiş verilere dayalı tahmin modellerinin geliştirilmesi gerekmektedir. Bu çalışmada, kredi notu veri seti üzerinde çeşitli makine öğrenmesi algoritmaları ile birlikte birliktelik kuralı çıkarımına dayalı Apriori algoritması kullanılarak tahmin modelleri oluşturulmuştur. Modelleme sürecinde veri madenciliği ve yapay zekâ tekniklerinden yararlanılmış; farklı sınıflandırma algoritmalarının başarı performansları 10 katlı çapraz doğrulama yöntemi ile doğruluk, hassasiyet (precision), hatırlama (recall) ve F1-skoru gibi metrikler üzerinden değerlendirilmiştir. İstatistiksel analizler (Wilcoxon ve paired t-testi) DNN modelinin Logistic Regression, Naive Bayes ve Random Forest modellerine göre anlamlı şekilde üstün olduğunu ortaya koyarken, MLP, SVM ve XGBoost modelleri ile benzer performanslar sergilediğini göstermiştir. Bu bulgu, DNN’nin özellikle karmaşık veri setlerinde güçlü bir tahmin modeli olduğunu desteklemektedir. Elde edilen sonuçlar, veri tabanlı tahmin yaklaşımlarının kredi risk analizinde etkinliğini göstermekte ve senaryoya uygun algoritma seçiminin önemine vurgu yapmaktadır. Ayrıca, kullanılan algoritmaların avantajları ve sınırlılıkları değerlendirilmiş; uygulama bağlamına göre en uygun yöntemin seçilmesinin kritik olduğu sonucuna varılmıştır. Bu çıktılar, kredi risk tahminine yönelik modelleme çalışmalarına katkı sağlamakta ve finansal kuruluşların karar destek sistemleri için yapay zekâ temelli çözümler geliştirmelerine rehberlik etmektedir.

Anahtar Kelimeler

• Veri Madenciliği
• Sınıflandırma Algoritmaları
• Kredi Skor Tahmin
• Birliktelik Analizi
• Hiperparametre Optimizasyonu

Predicting credit scores with data mining methods: performance comparison and analysis

Öz

Credit scoring prediction is critically important for financial institutions to effectively manage credit risk and ensure sustainable profitability. Accurate credit decisions require the development of predictive models based on historical data. In this study, predictive models were developed using various machine learning algorithms along with the Apriori algorithm based on association rule mining applied to a credit scoring dataset. The modeling process leveraged data mining and artificial intelligence techniques; the performance of different classification algorithms was evaluated using 10-fold cross-validation through metrics such as accuracy, precision, recall, and F1-score. Statistical analyses (Wilcoxon signed-rank test and paired t-test) revealed that the Deep Neural Network (DNN) model outperformed Logistic Regression, Naive Bayes, and Random Forest models significantly, while exhibiting similar performance to MLP, SVM, and XGBoost models. This finding supports the strength of DNN models, particularly on complex datasets. The results demonstrate the effectiveness of data-driven predictive approaches in credit risk analysis and emphasize the importance of selecting algorithms appropriate to the scenario. Furthermore, the advantages and limitations of the applied algorithms were evaluated, concluding that choosing the most suitable method according to the application context is critical. These findings contribute to credit risk prediction modeling efforts and provide guidance for financial institutions in developing AI-based decision support systems.

Anahtar Kelimeler

• Data Mining
• Classification Algorithms
• Credit Score Estimation
• Association Analysis
• Hyperparameter Tuning

Kaynakça

1. [1] Han J, Kamber M, Pei J. Data Mining: Concepts and Techniques. 3rd ed. Waltham, USA, Morgan Kaufmann Publishers, 2012.
2. [2] Tan PN, Steinbach M, Kumar V. Introduction to Data Mining. 2nd ed. Upper Saddle River, USA, Pearson Education, 2016.
3. [3] Provost F, Fawcett T. Data Science for Business: What You Need to Know about Data Mining and Data-Analytic Thinking. 1st ed. Sebastopol, USA, O'Reilly Media, 2013.
4. [4] Noman SM, Fadel YM, Henedak MT, Attia NA, Essam M, Elmaasarawii S, Fouad FA, Eltasawi EG, Al-Atabany W. “Leveraging survival analysis and machine learning for accurate prediction of breast cancer recurrence and metastasis”. Scientific Reports, 15(1), 3728, 2025.
5. [5] Duda RO, Hart PE, Stork DG. Pattern Classification. 2nd ed. New York, USA, Wiley, 2012.
6. [6] Hastie T, Tibshirani R, Friedman J. The Elements of Statistical Learning. 2nd ed. New York, USA, Springer, 2009.
7. [7] Rahman FMS, Islam MS. “Data clustering: Application and trends”. Artificial Intelligence Review, 56(1), 263-291, 2022.
8. [8] James G, Witten D, Hastie T, Tibshirani R. An Introduction to Statistical Learning. 2nd ed. New York, USA, Springer, 2021.

9. [9] Goodfellow I, Bengio Y, Courville A. Deep Learning. 1st ed. Cambridge, USA, MIT Press, 2016.
10. [10] Shi S, Tse R, Luo W, D’Addona S, Pau G. “Machine learning-driven credit risk: A systemic review”. Neural Computing and Applications, 34(17), 14327–14339, 2022.
11. [11] Moreira E, Moro S, Cortez P. “Deep learning and machine learning techniques for credit scoring: A review”. Communications in Computer and Information Science, 2069, 30–61, 2024.
12. [12] Huang CL, Chen MC, Wang CJ. "Credit scoring with a data mining approach based on support vector machines." Expert Systems with Applications, 33(4), 847-856, 2007.
13. [13] Xia Y, Liu C, Li Y, Liu N. "An ensemble credit scoring model based on a sequential ensemble system with XGBoost." Expert Systems with Applications, 125, 291-306, 2019.
14. [14] Li H, Wang Q, Zhao J. “Neural networks for credit risk assessment: Accuracy and training time”. Neural Computing and Applications, 32(5), 1423–1435, 2020.
15. [15] Oreski S, Oreski G. "Genetic algorithm-based heuristic for feature selection in credit risk assessment." Expert Systems with Applications, 41(4), 2052-2064, 2014.
16. [16] Zhou F, Huang T, Yang P. “Deep learning models for credit risk prediction: A comparative study”. Expert Systems with Applications, 215, 119526, 2023.
17. [17] Ghosh R, Kumar S, Roy S. “Logistic regression vs. other classifiers in credit scoring”. Journal of Risk and Financial Management, 13(7), 157, 2020.
18. [18] Zontul M, Polat H, Yıldırım T. “Customer credit rating estimation using machine learning methods”. Applied Soft Computing, 106, 107312, 2021.
19. [19] Arram A, Ayob M, Albadr MAA, Sulaiman A, Albashish D. “Comparing the effectiveness of machine learning and deep learning models in student credit scoring: A case study in Vietnam”. Applied Soft Computing, 142, 110839, 2023.
20. [20] Hussain A, Khan MA, Ahamed AKS, Kousarziya S. “Enhancing credit scoring models with artificial intelligence: A comparative study of traditional methods and AI-powered techniques”. Expert Systems with Applications, 230, 120099, 2023.
21. [21] Mhlanga D. “Machine learning approaches to credit risk: Comparative study of Islamic and conventional banks”. Journal of Risk and Financial Management, 14(12), 608, 2021.
22. [22] Sarıman G. “Veri madenciliğinde kümeleme teknikleri üzerine bir çalışma: K-Means ve K-Medoids kümeleme algoritmalarının karşılaştırılması”. Süleyman Demirel Üniversitesi Fen Bilimleri Dergisi, 15(3), 192-198, 2011.
23. [23] Thomas LC. Consumer Credit Models: Pricing, Profit and Portfolios. 2nd ed. Oxford, UK, Oxford University Press, 2020.
24. [24] Hair JF, Black WC, Babin BJ, Anderson RE, Tatham RL. Multivariate Data Analysis. 6th ed. Upper Saddle River, USA, Pearson Prentice Hall, 2006.
25. [25] Manning CD, Raghavan P, Schütze H. Introduction to Information Retrieval. 1st ed. Cambridge, UK, Cambridge University Press, 2008.
26. [26] Rokach L, Maimon O. Data Mining with Decision Trees: Theory and Applications. 2nd ed. Singapore, Singapore, World Scientific, 2014.
27. [27] Breiman L. “Random forests”. Machine Learning, 45(1), 5–32, 2001.
28. [28] Huang CL, Wang CJ. “A GA-based feature selection and parameters optimization for support vector machines”. Expert Systems with Applications, 31(2), 231–240, 2006.
29. [29] Chen T, Guestrin C. “XGBoost: A scalable tree boosting system”. 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, San Francisco, USA, 2016.
30. [30] Ruder S. “An overview of gradient descent optimization algorithms”. arXiv, 2016.
31. [31] Diehl PU, Cook M. “Unsupervised learning of digit recognition using spike-timing-dependent plasticity”. Frontiers in Computational Neuroscience, 9, 99, 2015.
32. [32] Powers DMW. “Evaluation: From precision, recall and F-measure to ROC, informedness, markedness and correlation”. Journal of Machine Learning Technologies, 2(1), 37–63, 2011.
33. [33] Agrawal R, Imieliński T, Swami A. “Mining association rules between sets of items in large databases”. ACM SIGMOD International Conference on Management of Data, Washington, USA, 1993.
34. [34] Zaki MJ, Meira W. Data Mining and Analysis: Fundamental Concepts and Algorithms. 2nd ed. Cambridge, UK, Cambridge University Press, 2020.
35. [35] Sahami Shirazi A, Norouzi F. “A review on association rule mining methods and applications”. Journal of Big Data, 8(1), 86, 2021.
36. [36] Kotsiantis SB, Zaharakis I, Pintelas P. “Supervised machine learning: A review of classification techniques”. Frontiers in Artificial Intelligence Applications, 160, 159–184, 2007.