COST-SENSITIVE EXPLAINABLE AI FOR CORPORATE BANKRUPTCY PREDICTION USING FINANCIAL RATIOS
Keywords:
Bankruptcy Prediction, Financial Distress, Explainable AI, Financial Ratios, Cost-Sensitive Learning, Business AnalyticsAbstract
The prediction of corporate bankruptcy is at the heart of financial risk management, but its infrequent occurrence and thelack of transparency in the models remain limiting its application in business and accounting practice. In response tothese issues, an explainable machine learning framework is constructed in a cost-sensitive way to predict bankruptcybased on structured financial ratios. The algorithm combines gradient boosting, where LightGBM will be used as themain classifier, class-weighted optimization, and systematic threshold calibration to solve asymmetric misclassificationrisk. Stratified cross-validation and minority-centric measures, especially Precision-Recall Area Under the Curve (PR-AUC), are used to measure model performance, in conjunction with F 1-score, recall and balanced accuracy. SHapleyAdditive exPlanations (SHAP) is used to measure global feature importance and nonlinear financial impact to make surethat the managerial interpretation is possible. Empirical results show strong discriminatory performance in the case ofextreme class imbalance, with balanced precision and recall, and reliably identifying liquidity and leverage ratios as thekey factors in the distress risk. The findings indicatethat the applicability of interpretable boosting models can beconsidered as precursory alert systems in credit screening, investment screening, and regulatory monitoring. Thecombination of explainable artificial intelligence with cost-sensitive learningin a single financial ratio model providesmethodologically sound and decision-relevant information to the modern business analytics and accounting research.
Downloads
References
Chen, T., & Guestrin, C. (2016, August). Xgboost: A scalable tree boosting system. InProceedings of the 22nd acm
sigkdd international conference on knowledge discovery and data mining(pp. 785-794).
Ribeiro, M. T., Singh, S., & Guestrin, C. (2016, August). " Why should i trust you?" Explaining the predictions of any
classifier. InProceedings of the 22nd ACM SIGKDD international conference on knowledge discovery and datamining(pp. 1135-1144).
Zięba, M., Tomczak, S. K., & Tomczak, J. M. (2016). Ensemble boosted trees with synthetic features generation in
application to bankruptcy prediction.Expert systems with applications,58, 93-101.
Barboza, F., Kimura, H., & Altman, E. (2017).Machine learning models and bankruptcy prediction.Expert systems
with applications,83, 405-417.
Ke, G., Meng, Q., Finley, T., Wang, T., Chen, W., Ma, W., ... & Liu, T. Y. (2017). Lightgbm: A highly efficientgradient
boosting decision tree.Advances in neural information processing systems,30.
Lundberg, S. M., & Lee, S. I. (2017).A unified approach to interpreting model predictions.Advances in neural
information processing systems,30.
Lin, T. Y., Goyal,P., Girshick, R., He, K., & Dollár, P. (2017). Focal loss for dense object detection. InProceedings of
the IEEE international conference on computer vision(pp. 2980-2988).
Boughorbel, S., Jarray, F., & El-Anbari, M. (2017). Optimal classifier for imbalanced data using Matthews Correlation
Coefficient metric.PloS one,12(6), e0177678.
Prokhorenkova, L., Gusev, G., Vorobev, A., Dorogush, A. V., & Gulin, A. (2018). CatBoost: unbiased boosting with
categorical features.Advances in neural information processing systems,31.
Hosaka, T. (2019). Bankruptcy prediction using imaged financial ratios and convolutional neural networks.Expert
systems with applications,117, 287-299.
Chicco, D., & Jurman, G. (2020). The advantages of the Matthews correlation coefficient (MCC) over F1 score and
accuracy in binary classification evaluation.BMC genomics,21(1), 6.
Yerashenia, N., Bolotov, A., Chan, D., & Pierantoni, G. (2020, June). Semantic data pre-processing for machine
learning based bankruptcy prediction computational model. In2020 IEEE 22nd Conference on Business Informatics
(CBI)(Vol. 1, pp. 66-75). IEEE.
Ben Jabeur, S., Stef, N., & Carmona, P. (2023).Bankruptcy prediction using the XGBoost algorithm and variable
importance feature engineering.Computational Economics,61(2), 715-741.
Kuizinienė, D., Krilavičius, T., Damaševičius, R., & Maskeliūnas, R. (2022). Systematic review of financial distress
identification using artificial intelligence methods.Applied Artificial Intelligence,36(1), 2138124.
Hacibedel, M. B., & Qu, R. (2022).Understanding and predicting systemic corporate distress: a machine-learning
approach. International Monetary Fund.
Zhong, J., & Wang, Z. (2022). Artificial intelligence techniques for financial distress prediction.AIMS
Mathematics,7(12), 20891-20908.
Chen, T. K., Liao, H. H., Chen, G. D., Kang, W. H., & Lin, Y. C. (2023).Bankruptcy prediction using machine learning
models with the text-based communicative value of annual reports.Expert Systems with Applications,233, 120714.18.Florek, P., & Zagdański, A. (2023). Benchmarking state-of-the-art gradient boosting algorithms for
classification.arXiv preprint arXiv:2305.17094.
Taiwanese Bankruptcy Prediction. (2020).UCI Machine Learning Repository. University of California, Irvine, School
of Information and Computer Sciences.https://archive.ics.uci.edu/dataset/572/taiwanese+bankruptcy+prediction
