M.N. Belozerov1, A.N. Smirnov2

1 National Research Technological University "MISIS" (Moscow, Russia)

2 SberBank of Russia PJSC, Headquarters (Moscow, Russia)

1mnbelozyorov@gmail.com, 2asmirnov889@yandex.ru

Problem statement. Modeling is used in business process management, providing hardware and software tools for optimization, forecasting, and data analysis with the ultimate goal of making strategically sound decisions. At the same time, the introduction of machine learning models into the production environment involves corresponding model risks that significantly affect the efficiency of the enterprise. In this regard, the task of managing model risks to minimize them is urgent.

The purpose of the study is to develop a general scheme for automatic management of model risks in a manufacturing enterprise.

Results. A systematic approach to model risk management (DMR) has been developed, in which the management system functions without the need for human intervention. The proposed solution makes it possible to quickly identify problems at the stage of industrial operation and update the model version in a timely manner.

Practical significance. The proposed approach to working with models allows solving problems related to the development and implementation of models in business processes. Such automation reduces the cost of the UMP process, minimizes the impact of the human factor and significantly reduces the time during which the model carries risks.

Belozerov M.N., Smirnov A.N. Systematic approach to the automatic management of model risks // Information-measuring and Control Systems. 2026. V. 24. № 3. P. 19−25. DOI: https://doi.org/10.18127/j20700814-202603-02

1. Nikitin N.A. Veroyatnostnye metody ucheta modelnykh riskov pri otsenke investitsii v tekhnologii iskusstvennogo intellekta. Innovatsionnoe razvitie ekonomiki. 2023. T. 2. S. 123−134. (in Russian)
2. Dyadyunov D.A. Mashinnoe obuchenie dlya risk-menedzhmenta v banke: vozmozhnosti i vyzovy. Vestnik nauki. 2025. T. 1. № 1 (82). S. 265−273. (in Russian)
3. Moiseev E. i dr. Metod otsenki IT-sostavlyayushchei modelnogo riska i ekonomicheskogo kapitala na ego pokrytie. Journal of Money and Finance. 2022. T. 81. № 3. S. 107−127. (in Russian)
4. Svistunova S.A., Muzalev S.V. Ispolzovanie mashinnogo obuchenie v protsesse risk-menedzhmenta predmetnykh riskov. Russian Journal of Management. 2021. T. 9. № 3. S. 126−130. (in Russian)
5. Navarro C.L.A. et al. Risk of bias in studies on prediction models developed using supervised machine learning techniques: systematic review. BMJ. 2021. T. 375.
6. Lin S.S. et al. Risk assessment and management of excavation system based on fuzzy set theory and machine learning methods. Automation in Construction. 2021. V. 122. P. 103490.
7. Baryannis G. et al. Supply chain risk management and artificial intelligence: state of the art and future research directions. International journal of production research. 2019. V. 57. № 7. P. 2179−2202.
8. Song L., Mittal P. Systematic evaluation of privacy risks of machine learning models. 30th USENIX Security Symposium (USENIX Security 21). 2021. P. 2615−2632.
9. Schröer C., Kruse F., Gómez J.M. A systematic literature review on applying CRISP-DM process model. Procedia Computer Science. 2021. V. 181. P. 526−534.
10. Peker S., Kart Ö. Transactional data-based customer segmentation applying CRISP-DM methodology: A systematic review. Journal of Data, Information and Management. 2023. V. 5. № 1. P. 1−21.
11. Brzozowska J. et al. Data engineering in CRISP-DM process production data-case study. Applied Computer Science. 2023. V. 19. № 3.
12. Saltz J.S. CRISP-DM for data science: strengths, weaknesses and potential next steps. IEEE International Conference on Big Data (Big Data). 2021. P. 2337−2344.