Multi-Level Classification of Audit Opinions Using Ensemble Learning Methods with Encrypted Financial Data

Yıl 2025, Cilt: 18 Sayı: 3, 269 - 282, 31.07.2025

Elif Nur Kucur , Burak Aktürk , Ensar Yilmaz , Tolga Büyüktanır , Kazım Yıldız

https://doi.org/10.17671/gazibtd.1708959

Öz

Independent audit reports play a crucial role in assessing the financial reliability of companies. Auditors base their opinions on the accuracy and consistency of financial statements and their underlying components. This study aims to automatically predict audit opinions by leveraging financial ratios derived from financial statements, as well as well-known financial risk scores such as Altman-Z, Springate, and Zmijewski. Classification was performed using XGBoost and Random Forest algorithms. Considering data privacy requirements, the modeling process was implemented using the Concrete ML library, which supports homomorphic encryption, thereby preserving the confidentiality of financial data. A hierarchical classification approach was adopted further to subdivide unqualified audit opinions into more detailed sub-classes, enhancing interpretability. Experimental results show that the proposed model achieves strong performance in terms of both accuracy and F1 score. The developed system is expected to serve as a predictive, systematic, and privacy-aware decision support tool for auditors and other stakeholders prior to the formal audit process.

Anahtar Kelimeler

— privacy-preserving machine learning , audit opinion classification , hierarchical classification , financial ratios , ensemble learning

Şifrelenmiş Finansal Veriler ile Ensemble Öğrenme Yöntemleri Kullanılarak Bağımsız Denetim Görüşlerinin Çok Düzeyli Sınıflandırılması

Öz

Bağımsız denetim raporları, şirketlerin finansal güvenilirliğini değerlendirmede kritik bir rol oynamaktadır. Denetçiler, görüşlerini finansal tabloların doğruluğu ve tutarlılığı ile bu tabloları oluşturan bileşenlere dayandırmaktadır. Bu çalışma, finansal tablolardan türetilen finansal oranlar ile Altman-Z, Springate ve Zmijewski gibi bilinen finansal risk skorlarını kullanarak denetim görüşlerini otomatik olarak sınıflandırmayı amaçlamaktadır. Sınıflandırma işlemi XGBoost ve Random Forest algoritmalarıyla gerçekleştirilmiştir. Veri gizliliği gereksinimleri göz önünde bulundurularak, modelleme süreci homomorfik şifrelemeyi destekleyen Concrete ML kütüphanesi kullanılarak yürütülmüş ve böylece finansal verilerin gizliliği korunmuştur. Nitelikli denetim görüşlerini daha ayrıntılı alt sınıflara ayırmak amacıyla hiyerarşik bir sınıflandırma yaklaşımı benimsenmiş, bu sayede yorumlanabilirlik artırılmıştır. Deneysel sonuçlar, önerilen modelin hem doğruluk hem de F1 skoru açısından güçlü bir performans sergilediğini göstermektedir. Geliştirilen sistemin, resmi denetim süreci öncesinde denetçilere ve diğer paydaşlara öngörüye dayalı, sistematik ve gizliliği koruyan bir karar destek mekanizması sunması beklenmektedir.

Anahtar Kelimeler

mahremiyet koruyan makine öğrenmesi , bağımsız denetim görüşü sınıflandırma , hiyerarşik sınıflandırma , finansal oranlar , topluluk öğrenmesi

Kaynakça

• A. Yaşar, E. Yakut, M. M. Gutnu, “Predicting qualified audit opinions using financial ratios: Evidence from the Istanbul Stock Exchange”, International Journal of Business and Social Science, 6(8), 57–67, 2015.
• K. H. Chan, K. Z. Lin, R. R. Wang, “Government ownership, Accounting-Based regulations, and the pursuit of favorable audit opinions: Evidence from China”, Auditing: A Journal of Practice & Theory, 31(4), 47–64, 2012.
• N. Dopuch, R. W. Holthausen, R. W. Leftwich, “Predicting audit qualifications with financial and market variables”, Accounting Review, 62(3), 431–454, 1987.
• G. Husain, D. Nasef, R. Jose, J. Mayer, M. Bekbolatova, T. Devine, M. Toma, “SMOTE vs. SMOTEENN: A study on the performance of resampling algorithms for addressing class imbalance in regression models”, Algorithms, 18(1), 37, 2025.
• E. Kirkos, C. Spathis, A. Nanopoulos, Y. Manolopoulos, “Identifying Qualified Auditors’ Opinions: A Data Mining Approach”, Journal of Emerging Technologies in Accounting, 4, 183–197, 2007.
• P. T. T. Oanh, D. N. Hung, V. T. T. Van, “Forecasting Audit Opinions on Financial Statements: Statistical Algorithm or Machine Learning”, Electronic Journal of Applied Statistical Analysis, 17(1), 133–152, 2024.
• A. Habib, “A meta-analysis of the determinants of modified audit opinion decisions”, Managerial Auditing Journal, 28(3), 184–216, 2013.
• S. M. Saif, M. Sarikhani, F. Ebrahimi, “Finding Rules for Audit Opinions Prediction Through Data Mining Methods”, European Online Journal of Natural and Social Sciences, 1(2), 28–36, 2012.
• J. R. Sánchez-Serrano, D. Alaminos, F. García-Lagos, “Predicting Audit Opinion in Consolidated Financial Statements with Artificial Neural Networks”, Mathematics, 8, 1288, 2020.
• D. D. Tan, “Forecasting Audit Opinions on Financial Statements: Statistical Algorithm or Machine Learning”, Engineering and Technology Journal, 2024.
• T. Chen, C. Guestrin, “XGBoost: A Scalable Tree Boosting System”, Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD 2016), San Francisco, CA, A.B.D., 785–794, 2016.
• Chang Yu, Yixin Jin, Qianwen Xing, Ye Zhang, Shaobo Guo, Shuchen Meng. Advanced User Credit Risk Prediction Model using LightGBM, XGBoost and Tabnet with SMOTEENN.arXiv preprint arXiv:2408.03497v3, 2024.
• L. Breiman, “Random Forests”, Machine Learning, 45(1), 5–32, 2001.
• N. Stanišić, T. Radojević, N. Stanić, “Predicting the type of auditor opinion: Statistics, machine learning, or a combination of the two?”, Machine Learning, 1–58, 2019.
• A. K. Nawaiseh, M. F. Abbod, T. Itagaki, “Financial Statement Audit using Support Vector Machines, Artificial Neural Networks and K-Nearest Neighbor: An Empirical Study of UK and Ireland”, International Journal of Simulation–Systems, Science & Technology, 21(2), 2020.
• M. El-Bannany, M. Sreedharan, A. M. Khedr, “A robust deep learning model for financial distress prediction”, International Journal of Advanced Computer Science and Applications, 11(2), 2020.
• M. Elhoseny, N. Metawa, G. Sztanó, I. M. El-Hasnony, “Deep learning-based model for financial distress prediction”, Annals of Operations Research, 345(2), 885–907, 2025.
• E. I. Altman, M. Iwanicz-Drozdowska, E. K. Laitinen, A. Suvas, “Financial distress prediction in an international context: A review and empirical analysis of Altman's Z-score model”, Journal of International Financial Management & Accounting, 28(2), 131–171, 2017.
• A. Yaşar, “Olumlu görüş dışındaki denetim görüşlerinin veri madenciliği yöntemleriyle tahminine ilişkin karar ve birliktelik kuralları”, Mali Çözüm Dergisi / Financial Analysis, 26(133), 2016.
• H. Zarei, H. Yazdifar, M. D. Ghaleno, R. Azhmaneh, “Predicting auditors’ opinions using financial ratios and non-financial metrics: Evidence from Iran”, Journal of Accounting in Emerging Economies, 10(3), 425–446, 2020.
• M. Brygała, T. Korol, “Personal bankruptcy prediction using machine learning techniques”, Economics and Business Review, 10(2), 118–142, 2024.
• J. P. Sánchez Ballesta, E. García-Meca, “Audit qualifications and corporate governance in Spanish listed firms”, Managerial Auditing Journal, 20(7), 725–738, 2005.
• G. Husain, D. Nasef, R. Jose, J. Mayer, M. Bekbolatova, T. Devine, M. Toma, “SMOTE vs. SMOTEENN: A study on the performance of resampling algorithms for addressing class imbalance in regression models”, Algorithms, 18(1), 37, 2025.
• N. V. Chawla, K. W. Bowyer, L. O. Hall, W. P. Kegelmeyer, “SMOTE: Synthetic minority over-sampling technique”, J. Artif. Intell. Res., 16, 321–357, 2002.
• B. Adiloğlu, B. Vuran, “A multicriterion decision support methodology for audit opinions: The case of audit reports of distressed firms in Turkey”, Int. Bus. Econ. Res. J., 10(12), 37–48, 2011.
• A. Sideras, K. Bougiatiotis, E. Zavitsanos, G. Paliouras, G. Vouros, “Bankruptcy Prediction: Data Augmentation, LLMs and the Need for Auditor's Opinion”, Proceedings of the 5th ACM International Conference on AI in Finance, Association for Computing Machinery, 453–460, November 2024.
• U. Gupta, GPT-InvestAR: Enhancing stock investment strategies through annual report analysis with large language models, arXiv preprint arXiv:2309.03079, 2023.
• Y. Huang, Z. Wang, C. Jiang, “Diagnosis with incomplete multi-view data: A variational deep financial distress prediction method”, Technol. Forecast. Soc. Change, 201, 1–12, 2024.
• A. Saeedi, “A high-dimensional approach to predicting audit opinions”, Applied Economics, 55(33), 3807–3832, 2023.
• H. Zarei, H. Yazdifar, M. D. Dahmarde Ghaleno, R. Azhmaneh, “Predicting auditors' opinions using financial ratios and non-financial metrics: evidence from Iran”, Journal of Accounting in Emerging Economies, 10(3), 425–446, 2020.
• M. Todorovic, N. Stanisic, M. Zivkovic, N. Bacanin, V. Simic, E. B. Tirkolaee, “Improving audit opinion prediction accuracy using metaheuristics-tuned XGBoost algorithm with interpretable results through SHAP value analysis”, Applied Soft Computing, 149, 110955, 2023.
• Agra Fintech, B. Aktürk, AgraResearchLab, T. Büyüktanır, “Audit opinions of Turkish Public Companies [Data Set]”, Kaggle, https://www.kaggle.com/datasets/agrafintech/financial-data-of-turkish-public-companies, 28.05.2025.
• A. Stoian, B. Chevallier-Mames, “Zama Concrete ML: Simplifying Homomorphic Encryption for Python Machine Learning”, Python.org, https://www.python.org/success-stories/zama-concrete-ml-simplifying-homomorphic-encryption-for-python-machine-learning/, 28.05.2025.
• F. T. Kristanti, M. Y. Febrianta, D. F. Salim, H. A. Riyadh, Y. Sagama, B. A. H. Beshr, “Advancing Financial Analytics: Integrating XGBoost, LSTM, and Random Forest Algorithms for Precision Forecasting of Corporate Financial Distress”, Journal of Infrastructure, Policy and Development, 8(8), 4972, 2024.
• T. K. Ho, “Random decision forests”, Proceedings of the 3rd International Conference on Document Analysis and Recognition, Lausanne, Switzerland, 1, 278–282, IEEE, New York, NY, USA, 1995.
• K. Tissaoui, T. Zaghdoudi, A. Hakimi, M. Nsaibi, “Do gas price and uncertainty indices forecast crude oil prices? Fresh evidence through XGBoost modeling”, Computational Economics, 62, 663–687, 2022.
• M. Imani, A. Beikmohammadi, H. R. Arabnia, “Comprehensive analysis of random Forest and XGBoost performance with SMOTE, ADASYN, and GNUS under varying imbalance levels”, Technologies, 13(3), 88, 2025.
• E. Yılmaz, T. Büyüktanır, D. Civelek, B. Aktürk, AgraResearchLab, “Refined audit opinions using bankruptcy algorithms” [Data Set], Kaggle, https://www.kaggle.com/datasets/ensryilmaz/refined-audit-opinions-using-bankruptcy-algorithms/data, 28.05.2025
• F. T. Kristanti ve V. Dhaniswara, The accuracy of artificial neural networks and logit models in predicting the companies’ financial distress, Journal of Technology Management and Innovation, 2023.
• S. B. Kotsiantis, Supervised machine learning: A review of classification techniques, Informatica, 2007.
• D. M. W. Powers, “Evaluation: From precision, recall and F-factor to ROC, informedness, markedness & correlation”, Journal of Machine Learning Technologies, 2011, 2,