Voting Ensemble Framework with SMOTE for Financial Fraud Detection in Highly Imbalanced Transaction Data

Based on the very unbalanced nature of transaction data and the increasing complexity of fraudulent actions, financial fraud detection is still a challenging area. This study proposed a comprehensive machine learning-based fraud detection framework incorporating transaction data preparation, feature construction, and synthetic minority oversampling for class distribution balancing, and comparative evaluation of multiple predictive models. A large-scale financial transaction dataset containing more than 6.3 million records was utilized to assess the effectiveness of traditional machine learning, ensemble learning, boosting, deep learning, hybrid deep learning, and transformer-based approaches. When compared to deep learning models and conventional machine learning techniques, the experimental results demonstrated that ensemble learning techniques performed the best across all evaluation metrics. The overall performance showed that the proposed Voting Ensemble + SMOTE approach achieved the best overall performance. Confusion matrix analysis also verified the efficacy of the model, in correctly recognizing a high percentage of fraudulent transactions with an acceptable false alarm rate. From the results, it has been observed that the combination of all engineered features, imbalance mitigation strategies and ensemble modelling yields significant enhancement in the fraud detection capability. The proposed framework can help financial institutions to improve their transaction monitoring, minimize financial losses and increase the security of financial operations. On-going research could involve developing real-time fraud detection algorithms, fraud analytics tools based on graphs, explainable AI methods, and adaptive learning models that can analyse evolving fraud trends in a dynamic financial landscape.  

[1]                K. H. Ahmed, S. Axelsson, Y. Li, and A. M. Sagheer, “A credit card fraud detection approach based on ensemble machine learning classifier with hybrid data sampling,” Mach. Learn. with Appl., vol. 20, p. 100675, 2025, doi: https://doi.org/10.1016/j.mlwa.2025.100675. [2]                M. Akouhar, M. Ouhssini, M. El Fatini, A. Abarda, and E. Agherrabi, “Dynamic oversampling-driven Kolmogorov–Arnold networks for credit card fraud detection: An ensemble approach to robust financial security,” Egypt. Informatics J., vol. 31, p. 100712, 2025, doi: https://doi.org/10.1016/j.eij.2025.100712. [3]                Shpachuk, Vitaliy, Olena Markova, and Bogdan Adamyk. "AI-driven financial fraud: key risks and legal protections for financial institutions." Journal of Banking Regulation 27.1 (2026): 6. [4]                Jin, Jing, and Yongqing Zhang. "The analysis of fraud detection in financial market under machine learning." Scientific Reports 15.1 (2025): 29959. [5]                N. Almusallam and J. Qayyum, “A Hybrid Feature Selection and Clustering-Based Ensemble Learning Approach for Real-Time Fraud Detection in Financial Transactions,” Comput. Mater. Contin., vol. 85, no. 2, pp. 3653–3687, 2025, doi: https://doi.org/10.32604/cmc.2025.067220. [6]                Khan, Talha Ahmed, et al. "Analysis and visualization of fraud detection patterns through data mining and classification using MLP and hybrid deep learning model." Egyptian Informatics Journal 32 (2025): 100829. [7]                R. K. Gupta et al., “Enhanced framework for credit card fraud detection using robust feature selection and a stacking ensemble model approach,” Results Eng., vol. 26, p. 105084, 2025, doi: https://doi.org/10.1016/j.rineng.2025.105084. [8]                C. A. Cedric and H. Ould-Slimane, “Towards effective and robust bank fraud detection thanks to machine learning,” Procedia Comput. Sci., vol. 265, pp. 25–32, 2025, doi: https://doi.org/10.1016/j.procs.2025.07.152. [9]                E. Karnavou, G. Cascavilla, G. Marcelino, and Z. Geradts, “I know you’re a fraud: Uncovering illicit activity in a Greek bank transactions with unsupervised learning,” Expert Syst. Appl., vol. 288, p. 128148, 2025, doi: https://doi.org/10.1016/j.eswa.2025.128148. [10]                 Y. Chen, C. Zhao, Y. Xu, C. Nie, and Y. Zhang, “Deep Learning in Financial Fraud Detection: Innovations, Challenges, and Applications,” Data Sci. Manag., 2025, doi: https://doi.org/10.1016/j.dsm.2025.08.002. [11]                 K. EL Hachimi, G. Orhanou, and L. Ballihi, “Credit card fraud detection model based on optimized long short-term memory : Imbalanced class solution and overfitting reduction,” Expert Syst. Appl., vol. 305, p. 130871, 2026, doi: https://doi.org/10.1016/j.eswa.2025.130871. [12]                 H. Fanai and H. Abbasimehr, “A novel combined approach based on deep Autoencoder and deep classifiers for credit card fraud detection,” Expert Syst. Appl., vol. 217, p. 119562, 2023, doi: https://doi.org/10.1016/j.eswa.2023.119562. [13]                 Z. Chen, H. Chen, Y. Liu, and Y. Shi, “Multi-label financial statement fraud detection based on long short-term memory and multilayer perceptron hybrid model,” Eng. Appl. Artif. Intell., vol. 170, p. 114188, 2026, doi: https://doi.org/10.1016/j.engappai.2026.114188. [14]                 J. Du, C. Hu, H. Zhao, C. Liu, H. Peng, and J. Zhao, “SCALE: Adaptive Heterogeneous Graph Prototype Learning for Open-Set Unknown Financial Fraud Detection,” Expert Syst. Appl., p. 132749, 2026, doi: https://doi.org/10.1016/j.eswa.2026.132749. [15]                 E. Kim et al., “Champion-challenger analysis for credit card fraud detection: Hybrid ensemble and deep learning,” Expert Syst. Appl., vol. 128, pp. 214–224, 2019, doi: https://doi.org/10.1016/j.eswa.2019.03.042. [16]                 Eedala, Sri Harsha. Financial Fraud Detection Dataset. Kaggle, 2024, https://www.kaggle.com/datasets/sriharshaeedala/financial-fraud-detection-dataset. Accessed 1 June 2026.