{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,1,13]],"date-time":"2026-01-13T12:14:35Z","timestamp":1768306475087,"version":"3.49.0"},"reference-count":43,"publisher":"MDPI AG","issue":"1","license":[{"start":{"date-parts":[[2026,1,11]],"date-time":"2026-01-11T00:00:00Z","timestamp":1768089600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation of China","doi-asserted-by":"crossref","award":["72071026"],"award-info":[{"award-number":["72071026"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"crossref"}]},{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation of China","doi-asserted-by":"crossref","award":["72271040"],"award-info":[{"award-number":["72271040"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"crossref"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Systems"],"abstract":"<jats:p>Accurate credit risk assessment for small firms and farmers is crucial for financial stability and inclusion; however, many models still rely on binary default labels, overlooking the continuum of borrower vulnerability. To address this, we propose Transformer\u2013LightGBM\u2013Stacked Logistic Regression (TL-StackLR), a hybrid stacking framework for multi-class loan default prediction. The framework combines three learners: a Feature Tokenizer Transformer (FT-Transformer) for feature interactions, LightGBM for non-linear pattern recognition, and a stacked LR meta-learner for calibrated probability fusion. We transform binary labels into three risk tiers, Low, Medium, and High, based on quantile-based stratification of default probabilities, aligning the model with real-world risk management. Evaluated on datasets from 3045 firms and 2044 farmers in China, TL-StackLR achieves state-of-the-art ROC-AUC scores of 0.986 (firms) and 0.972 (farmers), with superior calibration and discrimination across all risk classes, outperforming all standalone and partial-hybrid benchmarks. The framework provides SHapley Additive exPlanations (SHAP) interpretability, showing how key risk drivers, such as income, industry experience, and mortgage score for firms and loan purpose, Engel coefficient, and income for farmers, influence risk tiers. This transparency transforms TL-StackLR into a decision-support tool, enabling targeted interventions for inclusive lending, thus offering a practical foundation for equitable credit risk management.<\/jats:p>","DOI":"10.3390\/systems14010078","type":"journal-article","created":{"date-parts":[[2026,1,12]],"date-time":"2026-01-12T09:13:01Z","timestamp":1768209181000},"page":"78","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["From Binary Scores to Risk Tiers: An Interpretable Hybrid Stacking Model for Multi-Class Loan Default Prediction"],"prefix":"10.3390","volume":"14","author":[{"ORCID":"https:\/\/orcid.org\/0009-0005-1390-0117","authenticated-orcid":false,"given":"Ghazi","family":"Abbas","sequence":"first","affiliation":[{"name":"School of Economics and Management, Dalian University of Technology, Dalian 116024, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Zhou","family":"Ying","sequence":"additional","affiliation":[{"name":"School of Economics and Management, Dalian University of Technology, Dalian 116024, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-2343-1992","authenticated-orcid":false,"given":"Muzaffar","family":"Iqbal","sequence":"additional","affiliation":[{"name":"College of Management, Sichuan Agricultural University, Yaan 625014, China"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"1968","published-online":{"date-parts":[[2026,1,11]]},"reference":[{"key":"ref_1","first-page":"19","article-title":"The Impact of Weak Credit Covenants on Financial Stability and Economic Growth in Kenya","volume":"15","author":"Barngetuny","year":"2025","journal-title":"Int. J. Educ. Manag. Stud."},{"key":"ref_2","doi-asserted-by":"crossref","first-page":"102855","DOI":"10.1016\/j.jdeveco.2022.102855","article-title":"Group lending with covariate risk","volume":"157","author":"Ahlin","year":"2022","journal-title":"J. Dev. Econ."},{"key":"ref_3","doi-asserted-by":"crossref","unstructured":"Iwendi, C., Khan, S., Anajemba, J.H., Mittal, M., Alenezi, M., and Alazab, M. (2020). The use of ensemble models for multiple class and binary class classification for improving intrusion detection systems. Sensors, 20.","DOI":"10.3390\/s20092559"},{"key":"ref_4","doi-asserted-by":"crossref","first-page":"923","DOI":"10.1080\/01605682.2019.1705193","article-title":"Multi-class misclassification cost matrix for credit ratings in peer-to-peer lending","volume":"72","author":"Wang","year":"2020","journal-title":"J. Oper. Res. Soc."},{"key":"ref_5","doi-asserted-by":"crossref","first-page":"421","DOI":"10.1111\/1468-0300.00066","article-title":"The definition of the grading scales in banks\u2019 internal rating systems","volume":"30","author":"Foglia","year":"2001","journal-title":"Econ. Notes"},{"key":"ref_6","doi-asserted-by":"crossref","first-page":"7492","DOI":"10.1287\/mnsc.2023.4952","article-title":"Climate risk and capital structure","volume":"69","author":"Ginglinger","year":"2023","journal-title":"Manag. Sci."},{"key":"ref_7","doi-asserted-by":"crossref","first-page":"1063","DOI":"10.1057\/palgrave.jors.2601922","article-title":"Hurdle models of loan default","volume":"56","author":"Moffatt","year":"2005","journal-title":"J. Oper. Res. Soc."},{"key":"ref_8","doi-asserted-by":"crossref","first-page":"26","DOI":"10.1016\/j.omega.2018.02.001","article-title":"Banking credit worthiness: Evaluating the complex relationships","volume":"83","author":"Bai","year":"2019","journal-title":"Omega"},{"key":"ref_9","doi-asserted-by":"crossref","first-page":"102755","DOI":"10.1016\/j.irfa.2023.102755","article-title":"A two-stage credit scoring model based on random forest: Evidence from Chinese small firms","volume":"89","author":"Zhou","year":"2023","journal-title":"Int. Rev. Financ. Anal."},{"key":"ref_10","doi-asserted-by":"crossref","first-page":"3713","DOI":"10.1002\/ijfe.2346","article-title":"Leveraging random forest in micro-enterprises credit risk modelling for accuracy and interpretability","volume":"27","author":"Uddin","year":"2022","journal-title":"Int. J. Financ. Econ."},{"key":"ref_11","doi-asserted-by":"crossref","unstructured":"Zhao, Y., and Lin, D. (2023). Prediction of Micro- and Small-Sized Enterprise Default Risk Based on a Logistic Model: Evidence from a Bank of China. Sustainability, 15.","DOI":"10.3390\/su15054097"},{"key":"ref_12","doi-asserted-by":"crossref","unstructured":"Oliveira, N.A.d., and Basso, L.F.C. (2025). Explaining Corporate Ratings Transitions and Defaults Through Machine Learning. Algorithms, 18.","DOI":"10.3390\/a18100608"},{"key":"ref_13","doi-asserted-by":"crossref","unstructured":"Iqbal, M., Qazi, A., Ahmad, N., and Altaf, M. (2025). From blueprint to bit stream: Unveiling metaverse adoption challenges in construction supply chains of emerging economies\u2014A TOE\u2013ISM\u2013machine learning framework. Eng. Constr. Arch. Manag., 1\u201326.","DOI":"10.1108\/ECAM-08-2025-1320"},{"key":"ref_14","doi-asserted-by":"crossref","unstructured":"Zhang, X., Zhang, T., Hou, L., Liu, X., Guo, Z., Tian, Y., and Liu, Y. (2025). Data-Driven Loan Default Prediction: A Machine Learning Approach for Enhancing Business Process Management. Systems, 13.","DOI":"10.3390\/systems13070581"},{"key":"ref_15","doi-asserted-by":"crossref","unstructured":"Bacevicius, M., and Paulauskaite-Taraseviciene, A. (2023). Machine Learning Algorithms for Raw and Unbalanced Intrusion Detection Data in a Multi-Class Classification Problem. Appl. Sci., 13.","DOI":"10.3390\/app13127328"},{"key":"ref_16","doi-asserted-by":"crossref","first-page":"224","DOI":"10.3846\/16111699.2017.1280844","article-title":"Modeling credit approval data with neural networks: An experimental investigation and optimization","volume":"18","author":"Chi","year":"2017","journal-title":"J. Bus. Econ. Manag."},{"key":"ref_17","doi-asserted-by":"crossref","first-page":"84","DOI":"10.1016\/j.inffus.2021.11.011","article-title":"Tabular data: Deep learning is not all you need","volume":"81","author":"Armon","year":"2022","journal-title":"Inf. Fusion"},{"key":"ref_18","first-page":"253","article-title":"Optimizing FT-Transformer: Sparse Attention for Improved Performance and Interpretability","volume":"23","author":"Tokimasa","year":"2024","journal-title":"Ind. Eng. Manag. Syst."},{"key":"ref_19","doi-asserted-by":"crossref","first-page":"3197","DOI":"10.1007\/s10115-022-01756-8","article-title":"Interpretable deep learning: Interpretation, interpretability, trustworthiness, and beyond","volume":"64","author":"Li","year":"2022","journal-title":"Knowl. Inf. Syst."},{"key":"ref_20","doi-asserted-by":"crossref","first-page":"240","DOI":"10.1016\/j.ijforecast.2021.03.013","article-title":"Deep learning for modeling the collection rate for third-party buyers","volume":"38","author":"Nazemi","year":"2022","journal-title":"Int. J. Forecast."},{"key":"ref_21","doi-asserted-by":"crossref","first-page":"589","DOI":"10.1111\/j.1540-6261.1968.tb00843.x","article-title":"Financial ratios, discriminant analysis and the prediction of corporate bankruptcy","volume":"23","author":"Altman","year":"1968","journal-title":"J. Financ."},{"key":"ref_22","doi-asserted-by":"crossref","first-page":"247","DOI":"10.1007\/s10479-024-06134-x","article-title":"Prediction of bank credit worthiness through credit risk analysis: An explainable machine learning study","volume":"354","author":"Chang","year":"2024","journal-title":"Ann. Oper. Res."},{"key":"ref_23","doi-asserted-by":"crossref","first-page":"91","DOI":"10.1016\/j.dss.2017.08.001","article-title":"Solvency prediction for small and medium enterprises in banking","volume":"102","author":"Figini","year":"2017","journal-title":"Decis. Support Syst."},{"key":"ref_24","doi-asserted-by":"crossref","unstructured":"Chen, T. (2016). XGBoost: A Scalable Tree Boosting System, Cornell University.","DOI":"10.1145\/2939672.2939785"},{"key":"ref_25","doi-asserted-by":"crossref","first-page":"13","DOI":"10.1016\/S0165-0114(83)80002-5","article-title":"A theory of possibility for decision making","volume":"9","author":"Borisov","year":"1983","journal-title":"Fuzzy Sets Syst."},{"key":"ref_26","doi-asserted-by":"crossref","unstructured":"Baydili, \u0130.T., and Tasci, B. (2025). Predicting employee attrition: Xai-powered models for managerial decision-making. Systems, 13.","DOI":"10.3390\/systems13070583"},{"key":"ref_27","first-page":"18932","article-title":"Revisiting deep learning models for tabular data","volume":"34","author":"Gorishniy","year":"2021","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"ref_28","unstructured":"EBA (2021). Discussion Paper on Machine Learning for IRB Models, European Banking Authority."},{"key":"ref_29","doi-asserted-by":"crossref","unstructured":"Shaikh, T.A., Rasool, T., Verma, P., and Mir, W.A. (2024). A fundamental overview of ensemble deep learning models and applications: Systematic literature and state of the art. Ann. Oper. Res., 1\u201377.","DOI":"10.1007\/s10479-024-06444-0"},{"key":"ref_30","doi-asserted-by":"crossref","unstructured":"Tahir, T., Jahankhani, H., Tasleem, K., and Hassan, B. (2025). Cross-Project Multiclass Classification of EARS-Based Functional Requirements Utilizing Natural Language Processing, Machine Learning, and Deep Learning. Systems, 13.","DOI":"10.3390\/systems13070567"},{"key":"ref_31","first-page":"4765","article-title":"A unified approach to interpreting model predictions","volume":"30","author":"Lundberg","year":"2017","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"ref_32","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1186\/s40100-025-00444-9","article-title":"Exit without choice: Interpretable machine learning unlocks the structural drivers of smallholder dispossession in Pakistan","volume":"14","author":"Abbas","year":"2026","journal-title":"Agric. Food Econ."},{"key":"ref_33","unstructured":"Postal Savings Bank of China (PSBC), and Dalian University of Technology (DUT) (2014). Credit Risk Lending Decision and Evaluation Report for Farmers, Postal Savings Bank of China Co., Ltd."},{"key":"ref_34","doi-asserted-by":"crossref","unstructured":"Abbas, G., Ying, Z., and Ayoubi, M. (2025). Consensus-driven feature selection for transparent and robust loan default prediction. Sci. Rep.","DOI":"10.21203\/rs.3.rs-7673650\/v1"},{"key":"ref_35","doi-asserted-by":"crossref","first-page":"417","DOI":"10.1080\/07474938.2017.1318469","article-title":"Common threshold in quantile regressions with an application to pricing for reputation","volume":"38","author":"Su","year":"2019","journal-title":"Econ. Rev."},{"key":"ref_36","doi-asserted-by":"crossref","first-page":"7499","DOI":"10.1109\/TNNLS.2022.3229161","article-title":"Deep neural networks and tabular data: A survey","volume":"35","author":"Borisov","year":"2022","journal-title":"IEEE Trans. Neural Netw. Learn. Syst."},{"key":"ref_37","doi-asserted-by":"crossref","first-page":"1509","DOI":"10.1007\/s00521-024-10698-x","article-title":"Sparse attention is all you need for pre-training on tabular data","volume":"37","author":"Isomura","year":"2025","journal-title":"Neural Comput. Appl."},{"key":"ref_38","doi-asserted-by":"crossref","unstructured":"Zhu, S., Wu, H., Ngai, E.W.T., Ren, J., He, D., Ma, T., and Li, Y. (2024). A Financial Fraud Prediction Framework Based on Stacking Ensemble Learning. Systems, 12.","DOI":"10.3390\/systems12120588"},{"key":"ref_39","doi-asserted-by":"crossref","first-page":"875","DOI":"10.1142\/S0218488525400136","article-title":"Credal Classification through an Ensemble of Confidence-Aware TabTransformers and its Application to Fraud Detection","volume":"33","author":"Galiani","year":"2025","journal-title":"Int. J. Uncertain. Fuzziness Knowl.-Based Syst."},{"key":"ref_40","unstructured":"Dong, Q., Zhu, X., and Gong, S. (February, January 27). Single-label multi-class image classification by deep logistic regression. Proceedings of the AAAI Conference on Artificial Intelligence, Honolulu, HI, USA."},{"key":"ref_41","doi-asserted-by":"crossref","first-page":"837","DOI":"10.2307\/2531595","article-title":"Comparing the areas under two or more correlated receiver operating characteristic curves: A nonparametric approach","volume":"44","author":"DeLong","year":"1988","journal-title":"Biometrics"},{"key":"ref_42","doi-asserted-by":"crossref","unstructured":"Du, Z., and Lv, G. (2025). Can Digital Finance Unleash the Potential for Household Consumption? A Comparison Based on the Inconsistency Between Income and Consumption Classes. J. Theor. Appl. Electron. Commer. Res., 20.","DOI":"10.3390\/jtaer20040275"},{"key":"ref_43","doi-asserted-by":"crossref","first-page":"355","DOI":"10.3934\/DSFE.2025015","article-title":"Leveraging hybrid ensemble models in stock market prediction: A data-driven approach","volume":"5","author":"Kayit","year":"2025","journal-title":"Data Sci. Finance Econ."}],"container-title":["Systems"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/2079-8954\/14\/1\/78\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,1,13]],"date-time":"2026-01-13T05:10:54Z","timestamp":1768281054000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/2079-8954\/14\/1\/78"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,1,11]]},"references-count":43,"journal-issue":{"issue":"1","published-online":{"date-parts":[[2026,1]]}},"alternative-id":["systems14010078"],"URL":"https:\/\/doi.org\/10.3390\/systems14010078","relation":{},"ISSN":["2079-8954"],"issn-type":[{"value":"2079-8954","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,1,11]]}}}