APPLICATION OF ENSEMBLE LEARNING ALGORITHMS FOR FRAUD DETECTION IN BANKING TRANSACTIONS
DOI:
https://doi.org/10.35546/kntu2078-4481.2025.3.2.23Keywords:
banking fraud, classification, random forest, linear regression, decision tree, neural networksAbstract
This paper examines the use of ensemble machine learning methods for detecting and preventing banking fraud, which today represents one of the key threats to financial stability and the security of clients of financial institutions. The rapid development of digital technologies creates both new opportunities for optimizing financial processes and new challenges associated with increasingly sophisticated fraud schemes. In this regard, the task of detecting suspicious transactions requires high-tech and reliable solutions.The proposed approach involves the use of an ensemble model that combines the results of several machine learning algorithms, which helps to compensate for the weaknesses of individual models and provide more stable classification.Special attention is given to preliminary data preprocessing: normalization, class balancing, and the selection of the most informative features that directly influence model accuracy. One of the key requirements of the study is to reduce the number of fraudulent transactions that may be incorrectly classified as legitimate, since such cases not only cause financial losses for the bank but also harm its reputation and customer trust.Within the study, a comparative analysis of the ensemble method and single machine learning models was conducted, identifying the advantages and disadvantages of the proposed solution. The choice of this approach is also driven by its high practicality, compatibility with financial systems, and ease of integration. The ensemble method makes it possible to combine the strengths of simple models while reducing the impact of their weaknesses on the final result.Overall, the choice of software should depend on the project’s technical requirements, and to achieve the best results, different models and approaches should be analyzed. The results confirm that the use of ensemble methods increases classification accuracy and reduces the likelihood of false positives. This makes the proposed approach a promising tool for enhancing the protection of banking institutions against fraud and minimizing financial risks.
References
Almarshad F. A., Gashgari G. A., Alzahrani A. I. A. Generative adversarial networks-based novel approach for fraud detection for the European cardholders 2013 dataset. IEEE Access. Vol. 11. 2023. Pp. 107348–107368.
Hancock J. T., Bauder R. A., Wang H., Khoshgoftaar T. M. Explainable machine learning models for Medicare fraud detection. Journal of Big Data. Vol. 10, no. 1. 2023.
Alsayaydeh J. A. J., Aziz A., Rahman A. I. A. Development of programmable home security using GSM system for early prevention. ARPN Journal of Engineering and Applied Sciences. Vol. 16, no. 1. 2021. Pp. 88–97.
Fedorchenko I., Oliinyk A., Alsayaydeh J. A. J. Modified genetic algorithm to determine the location of the distribution power supply networks in the city. ARPN Journal of Engineering and Applied Sciences. Vol.15, no. 23. 2020. Pp. 2850–2867.
Shakhovska N., Liaskovsky D., Augousti A., Liaskovska S., Martyn Y. Design and deployment of data developer toolkit in cloud manufacturing environments. CEUR-WS. Vol. 3699. 2024. Pp. 47–56.
Fedorchenko V., Yeroshenko O., Shmatko O., Kolomiitsev O., Omarov M. (2024) Password hashing methods and algorithms on the.Net platform, Advanced Information Systems. Vol. 8, no. 4. 2024. Pp. 82–92.
Islam M. A., Uddin M. A., Aryal S., Stea G. An ensemble learning approach for anomaly detection in credit card data with imbalanced and overlapped classes. Journal of Information Security and Applications. Vol. 78. 2023.
Abdul Salam M., Fouad K. M., Elbably D. L., Elsayed S. M. Federated learning model for credit card fraud detection with data balancing techniques. Neural Computing and Applications. Vol. 36, no. 11. 2024. Pp. 6231–6256.
