Support for decision making on microcredit issuance based on the use of the information and analytical system

   The strategy for the development of financial technologies implemented by the Central Bank of the Russian Federation involves work in the direction of digitalization of various areas of activity of domestic financial institutions to increase the sustainability of their business, in particular, by introducing data-driven management methods and tools. The challenge of the study is to develop an effective way to support lending decisions in microfinance organizations through the use of data collection, storage and analytics methods and the creation of appropriate infrastructure. In this work, as a management support tool in a microfinance organization (MFI), it is proposed to use an information and analytical system (IAS), which includes four modules (data - analytics - management - presentation) and provides an increase in the efficiency of decisions on issuing microloans. Based on retrospective data on customer behavior, MFIs developed and tested credit risk forecasting models using machine learning algorithms (CatBoost, Logistic Regression). The gradient boosting model was found to show a higher predictive ability than the logistic regression model. Gradient boosting model predictions were calibrated to improve the interpretability of the model predictions. Also, for the purposes of interpreting the logic of the model, the sensitivity of forecasts to customer-related factors was analyzed using the SHAP method. The results of the analysis of the model showed that at the threshold of the probability of default τ = 0.22, a microfinance organization can receive significant savings from the use of a system that predicts the risks of non-repayment of loans by customers. A methodology for assessing the economic efficiency of using the information and analytical system is proposed. The information system tested in the study as a tool to support management in MFIs and the methodology for assessing the economic efficiency of its use create the prerequisites for the further development of the concept and methods of management based on data in domestic financial institutions.

1. Potapov A.G. Main Trends of the Current Stage of Microfinance Market Development // Bulletin of Eurasian Science. 2024. Vol. 16, No. s6.

2. Omokhoa J.E., Ojo A.T., Adeyemi K.S. Leveraging AI and Technology to Optimize Financial Management and Operations in Microfinance Institutions and SMEs. 2025. URL: https://www.researchgate.net/publication/387873377_Leveraging_AI_and_Technology_to_Optimize_Financial_Management_and_Operations_in_Microfinance_Institutions_and_SMEs (date of access: 19. 10. 2025).

3. Frolov Yu. V. Intelligent Systems and Managerial Decisions. Moscow : Moscow City Pedagogical University, 2000. 294 p.

4. National Project “Data Economy and Digital Transformation of the State”. URL: http://government.ru/rugovclassifier/923/about/ (date of access: 11. 10. 2025).

5. Main Directions for the Development of Financial Technologies for 2025–2027 / Bank of Russia. Moscow, 2024. 624 p. URL: https://www.cbr.ru/Content/Document/File/166399/onfintech_2025-27.pdf (date of access: 19. 10. 2025).

6. Credit Risk Prediction Using Machine Learning and Deep Learning: A Study on Credit Card Customers / V. Chang [et al.] // Risks. 2024. Vol. 12 (11). P. 174.

7. Vetlugin D.D. Performance Management of Entrepreneurial Microfinance Organizations Using Machine Learning Algorithms // Journal of Economics and Management (DVFU). 2024. Vol. 19, No. 2. Pp. 43–58.

8. Polyakov K.L., Zhukova L.V. Modeling the Probability of Credit Default among Microfinance Organization Clients (Case Study of One MFO) // HSE Economic Journal. 2019. Vol. 23, No. 4. Pp. 497–523.

9. Zolotaryuk A.V. Possibilities for Solving Problems of Microfinance Organizations Using Intelligent Machine Learning Methods // Bulletin of the Financial University. 2018. No. 4. Pp. 71–83.

10. Chub V.S. Comparative Analysis of Machine Learning Methods in Credit Risk Assessment // Information Technologies. Educational Resources and Technologies. 2023. No. 3 (44). Pp. 81–91.

11. Marikar S.M., Bandara H.M.N. An Analysis of Data Driven Decision-Making Capabilities of Managers in Banks// arXiv preprint. 2020. arXiv: 2007.01862. URL: https://arxiv.org/abs/2007.01862 (date of access: 26. 10. 2025).

12. Filimonova E.V., Gavrilov V.Yu. Calibration of Credit Risk Probabilities for Short-Term and Long-Term Loans Using Isotonic Regression // Reports of the Russian Academy of Sciences. Mathematics Series. 2023. Vol. 514, No. 2. Pp. 15–23.

13. Dorofeev V.S., Volosatova T.M. Ensembling Outlier Detection Methods in Training Dataset Preparation // Modeling, Optimization and Information Technologies. 2022. No. 10 (3). Pp. 6–13. doi: 10.26102/2310-6018/2022.38.3.013.

14. A survey on missing data in machine learning // Journal of Big Data. 2021. Vol. 8, No. 42. URL: https://journalofbigdata.springeropen.com/articles/10.1186/s40537-021-00516-9 (date of access: 26. 10. 2025).

15. Poslavskaya E., Korolev A. Encoding categorical data: Is there yet anything ‘hotter’ than one-hot encoding? // arXiv preprint. 2023. arXiv:2312.16930. URL: https://arxiv.org/abs/2312.16930 (date of access: 26. 10. 2025).

16. CatBoost: unbiased boosting with categorical features / L. Prokhorenkova, G. Gusev, A. Vorobyov, A.V. Dorogush, A. Gulin // Advances in Neural Information Processing Systems (NeurIPS). 2018. Vol. 31. Pр. 6639–6649.

17. Lundberg S.M., Lee S.-I. A Unified Approach to Interpreting Model Predictions // Advances in Neural Information Processing Systems. 2017. URL: https://papers.nips.cc/paper/7062-a-unified-approach-to-inter-preting-model-predictions (date of access: 24. 10. 2025).

18. Starkov V.V., Gorbatova S.S., Vodolaga V.I. Development of an ETL Process Based on Open-Source Technologies for Delivering Data to Consumers // Modeling and Data Analysis. 2023. Vol. 13, No. 2. Pp. 180–193.

19. Abdiraimov N.A., Alimzhanova L.M. Use of the ETL Process for Creating a Data Warehouse in Second-Tier Banks // Intelligent Technologies in Transport. 2023. No. S1 (35-1).

20. Identification and Interpretation of Significant Factors Influencing Client Defaults in Microfinance Institutions Using Machine Learning Methods / Yu.V. Frolov, E.S. Zhidkova, T.M. Bosenko, A.L. Zhidkov // Cybernetics and Control Theory in Systems : proceedings of 13^th Computer Science On-line Conference 2024, Zlin, Czech Republic, 01–31. 01. 2024. Vol. 1119-2. Cham : Springer, 2024. Pр. 150-157. doi: 10.1007/978-3-031-70300-3_10.

21. Python Software Foundation. Python Standard Library. URL: https://docs.python.org/3/library/ (date of access: 13. 10. 2025).

22. Frolov Yu.V., Bosenko T.M., Konopelko E.S. Statistical Study of Factors Affecting the Risk of Lending by Microfinance Institutions // Software Engineering Research in System Science : proceedings of 12^th Computer Science On-line Conference 2023, Zlin, Czech Republic, 01–30.04.2023. Vol. 722-1. Cham : Springer, 2023. Pр. 361-369. doi: 10.1007/978-3-031-35311-6_37.

23. Hand D.J., Till R.J. A Simple Generalisation of the Area Under the ROC Curve for Multiple Class Classification Problems // Machine Learning. 2001. Vol. 45, No. 2. Pр. 171–186.

24. Bannykh A.A., Letchikov A.V. Methodology for Assessing Borrower Credit Risk Using Credit Bureau Scoring // Bulletin of Udmurt University. Series “Economics and Law”. 2013. No. 4.