Keywords: predictive analytics, ensemble machine learning methods, decision making, insurance risk assessment, loss forecasting
UDC 004.8:368
DOI: 10.26102/2310-6018/2026.62.11.010
This article examines the development of a predictive-analytical system for assessing insurance risks. Traditional methods for assessing risk and the reliability of potential clients often fail to ensure calculation accuracy. Existing studies rely on small or synthetic datasets, leading to significantly inflated results. The article describes the use of large anonymized datasets within an insurance predictive-analytical system to improve loss prediction accuracy without increasing operational costs. The study aims to enhance the efficiency of insurance claim assessment by implementing a loss forecasting system that does not raise operational expenses. To achieve this, the authors proposed a decision-support system architecture based on predictive analytics for insurance, comprising seven layers – ranging from the secure ingestion of raw data to the final results interpretation module. Data security was ensured through SHA-256 encryption and the use of irreversible hash functions, which preclude the recovery of original personal information in the event of a data breach. Ensemble machine learning methods formed the mathematical foundation for processing large anonymized datasets within the system. Ridge regression was employed to smooth feature weights, while XGBoost was used to address data classification, regression, and ranking tasks.
1. Acciarini Ch., Cappa F., Boccardelli P., Oriani R. How can organizations leverage big data to innovate their business models? A systematic literature review. Technovation. 2023;123:102713. https://doi.org/10.1016/j.technovation.2023.102713
2. Dwivedi Y.K., Sharma A., Rana N.P., Giannakis M., Goel P., Dutot V. Evolution of artificial intelligence research in Technological Forecasting and Social Change: Research topics, trends, and future directions. Technological Forecasting and Social Change. 2023;192:122579. https://doi.org/10.1016/j.techfore.2023.122579
3. Jaiswal R., Gupta Sh., Tiwari A.K. Big data and machine learning-based decision support system to reshape the vaticination of insurance claims. Technological Forecasting and Social Change. 2024;209:123829. https://doi.org/10.1016/j.techfore.2024.123829
4. Kashirskaya L.V., Turgaeva A.A. Internal control in insurance companies. Easy and reliable. Moscow: Rus-science; 2021. 355 p. (In Russ.).
5. Chen T., Guestrin C. XGBoost: A scalable tree boosting system. In: Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD 2016), 13–17 August 2016, San Francisco, CA, USA. New York: ACM; 2016. P. 785–794. https://doi.org/10.1145/2939672.2939785
6. Ovsyannikov A.V., Barashko O.G. Evaluation of the effectiveness of software mathematical tools for technical support of decision-making. In: Organizational and Technical Management in Intersectoral Complexes: Proceedings of the International Scientific and Technical Conference, 28–29 October 2004, Minsk, Belarus. Minsk: Belarusian State Technological University; 2004. P. 244–245. (In Russ.).
7. Ziatdinova Z.L. Mathematics of insurance: calculation of insurance premiums. In: Problems of Small and Medium-Sized Business Development in Agriculture in the Context of the Digital Economy: Proceedings of the III International Scientific and Practical Conference: Volume 2, 20–21 May 2025, Kazan, Russia. Kazan: Kazan State Agrarian University; 2025. P. 98–103. (In Russ.).
8. Kondrashov A.A., Kurashov M.M., Loskutova E.E. Use of medical information systems data for developing personalized pharmaceutical insurance products: architectural modeling. Vestnik of the Smolensk State Medical Academy. 2025;24(1):253–261. (In Russ.). https://doi.org/10.37903/vsgma.2025.1.36
9. Stepanova M.N., Gavrilenko Zh.A. Assessment of decision support tools related to the formation of a sound household insurance portfolio. Journal of Applied Research. 2023;(11):49–54. (In Russ.).
10. Kuvayskova Yu.E. Application of ensemble machine learning methods for predicting the technical state of an object. Izvestia of Samara Scientific Center of the RAS. 2021;23(1):111–114. (In Russ.). https://doi.org/10.37313/1990-5378-2021-23-1-111-114
11. Marnadze E.I. Digitalization of underwriting: a review of scoring systems and machine learning methods in personal insurance. Industrial Economics. 2025;(S2):125–132. (In Russ.). https://doi.org/10.47576/2949-1886.2025.9.9.019
Keywords: predictive analytics, ensemble machine learning methods, decision making, insurance risk assessment, loss forecasting
For citation: Kalach A.V., Matytsina I.A., Zolotaryov Y.V., Zdolnik V.V. Decision support in a predictive-analytic insurance system. Modeling, Optimization and Information Technology. 2026;14(11). URL: https://moitvivt.ru/ru/journal/article?id=2665 DOI: 10.26102/2310-6018/2026.62.11.010 (In Russ).
© Kalach A.V., Matytsina I.A., Zolotaryov Y.V., Zdolnik V.V. Статья опубликована на условиях лицензии Creative Commons Attribution-NonCommercial 4.0 International (CC BY-NS 4.0)Received 10.09.2026
Revised 09.11.2026
Accepted 18.11.2026