Integrated data-driven credit default prediction in Uganda using machine learning models

Authors

Keywords:

SVM, XGBoost, Credit defaulter, Prediction model, Financial data

Abstract

The prediction of credit facility defaulters is quite a challenge in Uganda, particularly for those without a formal banking history. Existing prediction models cater for prediction using conventional banking records which is not sufficient. The use of integrated data to cater for the unbanked population is required to further enhance financial inclusivity and stability in Uganda’s financial landscape. This study therefore aims at filling this gap by using machine learning techniques on a rich blend of financial data, including mobile money and Fintech (Financial Technology) services, as well as traditional banking records. Several machine learning algorithms used for loan default prediction were compared, such as Random Forest, Logistic Regression, Support Vector Machines (SVM), and Extreme Gradient Boosting (XGBoost). Random Forest Model showed 96.66% accuracy, 79.65% recall, 96.52% precision and 0.85 AUC. XGBoost model was found to have an accuracy of 95.23%; recall, 73.32%; precision, 94.11%; and Area Under the Curve (AUC) of 0.81. However, Random Forest performed best by all metrics with XGBoost following slightly. Logistic Regression showed 89.53% accuracy but had a very low recall at 43.24% and precision at 66.59%. SVM performed averagely with 93.21% accuracy and 62.80% recall all falling below that of XGBoost and Random Forest. Thus, the study revealed the significance of machine learning models like Random Forest and XGBoost for credit scoring prediction. Overall, it will improve the ability of institutions and policymakers to identify potential default borrowers so as to mitigate loan default rates and ensure economic growth in underserved communities through more accurate and inclusive credit evaluation tools.

Dimensions

[1] M. G. Kasim & T. Matsumoto, “Mobile money, remittances, and household welfare: Panel evidence from rural Uganda”, World Development 79 (2016) 127. https://doi.org/10.1016/j.worlddev.2015.11.006.

[2] T. Suri & W. Jack, “The long-run poverty and gender impacts of mobile money”, Science 354 (2016) 1288. https://doi.org/10.1126/science.aah5309.

[3] A. Demirguc-Kunt, L. Klapper, D. Singer & S. Ansar, “The global findex database 2017: measuring financial inclusion and the fintech revolution”, World Bank Publications 1 (2018) 126033. http://documents.worldbank.org/curated/en/332881525873182837.

[4] Zofi Cash, “Ugandan fintechs in measuring credit worthiness”, 2023. [Online]. https://furtherafrica.com/2023/06/08/uganda-fintech-startup-zofi-cash-receives-us1m-pre-seed/.

[5] P. Mukiibi, P. Mugambe & J. Kampumure, “The effect of credit reference bureau information sharing on credit assessment in financial institutions in Uganda”, Ugandan Journal of Management and Public Policy Studies 17 39. https://ujmpps.umi.ac.ug/index.php/ujmpps/article/view/47.

[6] M. Andrianaivo & K. Kpodar, “ICT financial inclusion, and growth: evidence from African countries”, 2011. [Online]. https://www.imf.org/external/pubs/ft/wp/2011/wp1173.pdf.

[7] A. Irumba & Z. Enock, ‘Role of financial literacy in decreasing loan default rates among micro-borrowers in Uganda”, Metropolitan Journal Of Academic Multidisciplinary Research 3 (2024) 200. https://www.researchgate.net/publication/384323510_Role_Of_Financial_Literacy_In_Decreasing_Loan_Default_Rates_Among_Micro-Borrowers_In_Uganda.

[8] Monitor, “Rein in growing loan default rate”, 2023. [Online]. https://www.monitor.co.ug/uganda/oped/editorial/rein-in-growing-loan-default-rate-4397274.

[9] B. Ssekiziyivu, R. Mwesigwa, M. Joseph & I. Nkote Nabeta, “Credit allocation, risk management and loan portfolio performance of MFIs—A case of Ugandan firms”, Cogent business & management 4 (2017) 1374921. http://dx.doi.org/10.1080/23311975.2017.1374921.

[10] E. Nzibonera & I. Waggumbulizi, “Loans and growth of small-scale enterprises in Uganda: a case study of Kampala Central business area”, African Journal of Business Management 14 (2020) 159. http://dx.doi.org/10.5897/AJBM2020.8985.

[11] Y. He, Y. Jian, T. Liu & H. Xue, “Optimization of machine learning models for prediction of personal loan default rate”, International Conference on Bigdata Blockchain and Economy Management 2022 (2022) 270. https://doi.org/10.2991/978-94-6463-030-5_29.

[12] X. Zhu, Q. Chu, X. Song, P. Hu & L. Peng, “Explainable prediction of loan default based on machine learning models” Data science and management 6 (2023) 123. https://doi.org/10.1016/j.dsm.2023.04.003.

[13] A. Kozina, Ł. Kuzmi´ nski, M. Nadolny, K. Miałkowska, P. Tutak, J. Janus´ & R. Krol, “The default of leasing contracts prediction using machine´ learning”, Procedia Computer Science 225 (2023) 424. http://dx.doi.org/10.1016/j.procs.2023.10.027.

[14] Tianyi Xu, “Comparative analysis of machine learning algorithms for consumer credit risk assessment”, International Conference on Computer Engineering, Information Science & Application Technology 4 (2024) 60. https://doi.org/10.62051/r1m3pg16.

[15] H. Suresh & J. Guttag, “Understanding potential sources of harm throughout the machine learning life cycle”, MIT Case Studies in Social and Ethical Responsibilities of Computing [Preprint], 2021. https://doi.org/10.21428/2c646de5.c16a07bb.

[16] N. Suhadolnik, J. Ueyama & S. Da Silva, “Machine learning for enhanced credit risk assessment: An empirical approach”, Journal of Risk and Financial Management 16 (2023) 496. https://doi.org/10.3390/jrfm16120496.

[17] K. Naveen, N. Pavitha & K. Ashutosh, “Machine learning for credit default prediction in SMES: a Study from emerging economy, Accountancy Business and the Public Interest 7 (2021) 54. https://abpi.uk/wp-content/uploads/2024/07/04SP2406.pdf.

[18] J. Xu, Z. Lu & Y. Xie, “Loan default prediction of Chinese P2P market: a machine learning methodology”, Scientific Reports 11 (2021) 18759. https://doi.org/10.1038/s41598-021-98361-6.

[19] H. J. Bature, D. D. Wisdom, T. T. Dufuwa & I. O. Ayetuoma, “Credit default prediction system using machine learning”, Computology: Journal of Applied Computer Science and Intelligent Technologies 3 (2023) 51. https://doi.org/10.17492/computology.v3i1.2305.

[20] R. Hlongwane, K. Ramaboa & W. Mongwe, “Enhancing credit scoring accuracy with a comprehensive evaluation of alternative data”, 19 (2024) e0303566. https://doi.org/10.1371/journal.pone.0303566.

[21] E. G. Dada, A. I. Birma & A. A. Gora, “Ensemble machine learning algorithm for cost-effective and timely detection of diabetes in Maiduguri, Borno State”, Journal of the Nigerian Society of Physical Sciences 6 (2024) 2175. https://journal.nsps.org.ng/index.php/jnsps/article/view/2175/345.

[22] E. Idongesit, U. L. Chinedu & A. E. Ibor, ”A stacked ensemble approach with resampling techniques for highly effective fraud detection in imbalanced datasets”, Journal of the Nigerian Society of Physical Sciences 7 (2025) 206. https://journal.nsps.org.ng/index.php/jnsps/article/view/2066/366.

[23] S. Lundberg & S. I. Lee, “A unified approach to interpreting model predictions”, presented at Conference on Neural Information Processing Systems, California, U.S., 4th December, 2017. http://dx.doi.org/10.48550/arXiv.1705.07874.

[24] L. Noriega, M. Zhang & J. Fernandez, “Comparative analysis of machine learning algorithms in default prediction: Decision trees, random forests, and support vector machines”, Journal of Credit Risk Analytics 34 (2023) 98. http://dx.doi.org/10.13140/RG.2.2.31652.14725.

[25] I. Steinwart & A. Christmann, “Support Vector Machines”, in Support Vector Machines, Springer, Berlin, Germany, 2008, pp. 164–201. https://link.springer.com/book/10.1007/978-0-387-77242-4.

Published

2026-02-01

How to Cite

Integrated data-driven credit default prediction in Uganda using machine learning models. (2026). Journal of the Nigerian Society of Physical Sciences, 8(1), 2649. https://doi.org/10.46481/jnsps.2026.2649

Issue

Section

Computer Science

How to Cite

Integrated data-driven credit default prediction in Uganda using machine learning models. (2026). Journal of the Nigerian Society of Physical Sciences, 8(1), 2649. https://doi.org/10.46481/jnsps.2026.2649