Comparative Analysis of the Predictive Risk Assessment Modeling Technique Using Artificial Intelligence
Tolkynbekova A. Koishiyeva D. Bissembayev A. Mukhammejanova D. Mukasheva A. Kang J.W.
September 2025Korean Institute of Electrical Engineers
Journal of Electrical Engineering and Technology
2025#20Issue 64509 - 4526 pp.
This study is aimed at conducting a comparative evaluation of methods of machine learning to classify and predict the effectiveness in the dynamically developing electronic financial sector. Traditional methods for assessing bank efficiency and sustainability often involve subjective judgments and resource-intensive processes, which limits their adaptability and accuracy. The study carefully analyzed six artificial intelligence models: random forest, support vector machine, logistic regression, K-nearest neighbors, Naive Bayes, and gradient boosting. The classification capabilities of these models were estimated by categorizing the performance into high-performing, medium-performing, and low-performing groups using quantitative metrics such as accuracy, recall, precision and F1-score. The dataset, including key financial indicators which are assets, liabilities, return on assets, return on equity, non-performing loans, liquidity ratios, and risk-weighted assets, net income was collected from 21 banks over 6.5 years. The analysis showed that the support vector machine and logistic regression models achieved excellent forecasting accuracy exceeding 90%, demonstrating their robustness in identifying different performance classes. Random forest and gradient boosting models have achieved high results, especially in identifying between medium and top financial institutions. However, the K-nearest neighbors and Naive Bayes models faced difficulties when addressing the intricacies and inconsistencies of financial data. The findings of the research highlight the revolutionary capabilities of machine learning in developing automated predictive analysis systems in the banking sector, providing a scalable, unbiased, and data-driven alternative to traditional methods. This approach contributes to the development of algorithmic solutions for digital banking, expanding the use of artificial intelligence in the financial sector and creating the basis for the implementation of complex analytical platforms. This research lays the groundwork for deploying sophisticated analytics systems in the banking sector, ultimately leading to enhanced operational effectiveness.
Algorithm performance analysis , Artificial intelligence , Classification algorithms , Predictive modeling , Risk Assessment techniques
Text of the article Перейти на текст статьи
School of Information Technology and Engineering, Kazakh-British Technical University, Tole Bi Street 59, Almaty, Kazakhstan
Department of Artificial Intelligence and Big Data, Al-Farabi Kazakh National University, Almaty, Kazakhstan
Department of Transportation System Engineering, Korea National University of Transportation, Gyeonggi-do, Uiwang-si, South Korea
School of Information Technology and Engineering
Department of Artificial Intelligence and Big Data
Department of Transportation System Engineering
10 лет помогаем публиковать статьи Международный издатель
Книга Публикация научной статьи Волощук 2026 Book Publication of a scientific article 2026