Credit Card Fraud Detection Using Stacking Ensemble of Deep Learning Model

Abstract

Financial protection through credit card fraud detection demands sophisticated techniques to properly identify fraudulent payments among all transactions. Modern fraudulent activities create substantial hurdles for existing detection systems because fraudulent transactions remain sparse in relation to ordinary transactions. This research paper puts forth an improved fraud detection method by implementing a hybrid SMOTEENN resampling approach within a stacking ensemble system. A stacking ensemble model integrates Long Short-Term Memory (LSTM) networks together with Random Forest as its base learners, while utilizing a Multi-Layer Perceptron (MLP) to serve as the meta-learning model. The proposed detection system produces enhanced results through time pattern analysis and efficient treatment of unbalanced data distribution. The experimental trials prove the system\'s resilience and its result exceeds traditional machine learning models for reliable fraud act detection.

Introduction

1. Problem Statement

Credit card fraud detection is increasingly critical due to the rise in online and wireless payment systems. Traditional rule-based and statistical methods struggle with:

• Adaptability to evolving fraud tactics

• Imbalanced datasets (few fraudulent vs. many legitimate transactions)

• Non-stationary fraud patterns and noisy data

2. Proposed Solution

The paper presents a hybrid deep learning system that combines:

• SMOTE-ENN (Synthetic Minority Oversampling Technique with Edited Nearest Neighbors) for handling class imbalance and noise

• Ensemble learning integrating:

  • LSTM (Long Short-Term Memory) for sequential pattern recognition

  • Random Forest for robust classification

  • MLP (Multi-Layer Perceptron) as a meta-learner to aggregate predictions

3. Methodology Overview

• Data Preprocessing: Feature extraction, class separation, 80-20 train-test split

• Resampling: SMOTE generates synthetic fraud samples; ENN cleans noisy data

• Standardization: Normalizes feature values using StandardScaler

• Model Training:

  • LSTM for time-sequenced pattern detection

  • Random Forest for high accuracy with structured data

  • MLP to combine outputs of LSTM and Random Forest in a stacking ensemble

• Model Evaluation: Metrics include Accuracy, Precision, Recall, F1-Score

• Deployment: Flask-based web app interface for user transaction prediction; models are saved and loaded using joblib and TensorFlow

4. Literature Survey

Summarizes related works on:

• Class imbalance handling (e.g., SMOTE, SMOTE-ENN)

• Deep learning and ensemble techniques in fraud detection

• Synthetic data generation (GANs)

• Hybrid and federated learning approaches

5. Results

• All models achieved perfect precision, recall, and F1-score

• Ensemble model outperformed individual models slightly in accuracy

Conclusion

The detection capabilities exhibit potential improvement through different strategies which include deep learning along with ensemble learning approaches. The proposed hybrid stacking ensemble utilizes LSTM and Random Forest learners together with SMOTE-ENN resampling methods to handle the problems of imbalanced datasets. These experimental results show that the ensemble model proves effective at minimizing false positives along with false negatives which present the most-detrimental effects in fraud detection practices. The work needs additional research to enhance scalability and resilience as well as adaptability when applied to real-life scenarios. Model accuracy requires parallel optimization to computational efficiency because different algorithms perform differently between training time and running time measurements. This study recommends continued research to develop better fraud detection tactics for financial security against illegal transactions because future efforts will concentrate on maximizing model execution while enhancing their detection capabilities.

References

[1] Mienye, I. D., & Sun, Y. (2023). A deep learning ensemble with data resampling for credit card fraud detection. Ieee Access, 11, 30628-30638. [2] Soleymanzadeh, R., Aljasim, M., Qadeer, M. W., & Kashef, R. (2022). Cyberattack and fraud detection using ensemble stacking. AI, 3(1), 22-36. [3] 3.Cherif, A., Badhib, A., Ammar, H., Alshehri, S., Kalkatawi, M., & Imine, A. (2023). Credit card fraud detection in the era of disruptive technologies: A systematic review. Journal of King Saud UniversityComputer and Information Sciences, 35(1), 145-174. [4] Chugh, B., Garg, P., & Dwivedi, K. (2024). A Comprehensive Ensemble Approach Using Blending and Stacking for Credit Card Fraud Detection. International Journal on Smart & Sustainable Intelligent Computing, 1(1), 19-33. [5] Ileberi, E., & Sun, Y. (2024). A Hybrid Deep Learning Ensemble Model for Credit Card Fraud Detection. IEEE Access. [6] Khalid, A. R., Owoh, N., Uthmani, O., Ashawa, M., Osamor, J., & Adejoh, J. (2024). Enhancing credit card fraud detection: an ensemble machine learning approach. Big Data and Cognitive Computing, 8(1), 6 [7] Abbasi, M., & Shah, M. A. (2022, June). Credit card fraud detecting using machine learning classifiers in stacking ensemble technique. In IET Conference Proceedings CP801 (Vol. 2022, No. 8, pp. 76-81). Stevenage, UK: The Institution of Engineering and Technology. [8] Khan, F. N., Khan, A. H., & Israt, L. (2020, June). Credit card fraud prediction and classification using deep neural network and ensemble learning. In 2020 IEEE Region 10 Symposium (TENSYMP) (pp. 114-119). IEEE. [9] Esenogho, E., Mienye, I. D., Swart, T. G., Aruleba, K., & Obaido, G. (2022). A neural network ensemble with feature engineering for improved credit card fraud detection. IEEE access, 10, 16400-16407. [10] Eteng, I. E., Chinedu, U. L., & Ibor, A. E. (2025). A stacked ensemble approach with resampling techniques for highly effective fraud detection in imbalanced datasets. Journal of the Nigerian Society of Physical Sciences, 2066-2066. [11] KohilaKanagalakshmi, T., Janakiraman, S., & Suresh, S. (2024, January). Stacked Ensemble-Based Machine Learning Approach for Anomaly Detection in Credit Card Transactions. In The International Conference on Universal Threats in Expert Applications and Solutions (pp. 211-221). Singapore: Springer Nature Singapore. [12] Jyoti, E., Jindal, C., Singh, N., & Saini, A. (2024, November). Improving Credit Card Fraud Detection Through Stacking Ensemble Models and SMOTE-ENN for Imbalanced Datasets. In 2024 4th International Conference on Technological Advancements in Computational Sciences (ICTACS) (pp. 1642-1649). IEEE. [13] Kurien, K. L., & Chikkamannur, A. A. (2024). A stacking ensemble for credit card fraud detection using SMOTE technique. International Journal of Engineering Systems Modelling and Simulation, 15(6), 284-290. [14] Bakhtiari, S., Nasiri, Z., & Vahidi, J. (2023). Credit card fraud detection using ensemble data mining methods. Multimedia Tools and Applications, 82(19), 29057-29075. [15] Sobanadevi, V., & Ravi, G. (2021). Handling data imbalance using a heterogeneous bagging-based stacked ensemble (HBSE) for credit card fraud detection. In Intelligence in Big Data Technologies—BeyondtheHype: Proceedings of ICBDCC 2019 (pp. 517525). Springer Singapore. [16] Damanik, N., & Liu, C. M. (2025). Advanced Fraud Detection: Leveraging KSMOTEENN and Stacking Ensemble to Tackle Data Imbalance and Extract Insights. IEEE Access. [17] Nayyer, N., Javaid, N., Akbar, M., Aldegheishem, A., Alrajeh, N., & Jamil, M. (2023). A new framework for fraud detection in bitcoin transactions through ensemble stacking model in smart cities. IEEE Access, 11, 90916-90938. [18] Menon, P. P., & Sachdeva, A. (2024, May). Fraud Shield: Credit Card Fraud Detection with Ensemble and Deep Learning. In 2024 4th International Conference on Pervasive Computing and Social Networking (ICPCSN) (pp. 224-230). IEEE. [19] Veigas, K. C., Regulagadda, D. S., & Kokatnoor, S. A. (2021). Optimized stacking ensemble (OSE) for credit card fraud detection using synthetic minority oversampling model. Indian Journal of Science and Technology, 14(32), 2607-2615. [20] Zhou, X., Gururajan, R., Chan, K. C., & Alsodi, O. Enhancing Credit Card Fraud Detection with a Stacking-Based Hybrid Machine Learning Approach. [21] Mehnaz, F., Chaitra, K. M., & Basthikodi, M. (2024, November). A Fusion Approach for Credit Card Fraud Detection: A Stacked Ensemble Framework. In 2024 International Conference on Computing, Semiconductor, Mechatronics, Intelligent Systems and Communications (COSMIC) (pp. 157-162). IEEE. [22] Laveti, R. N., Mane, A. A., & Pal, S. N. (2021, April). Dynamic stacked ensemble with entropy based undersampling for the detection of fraudulent transactions. In 2021 6th International Conference for Convergence in Technology (I2CT) (pp. 17). IEEE. [23] Mim, M. A., Majadi, N., & Mazumder, P. (2024). A soft voting ensemble learning approach for credit card fraud detection. Heliyon, 10(3). [24] Baabdullah, T., Rawat, D. B., Liu, C., & Alzahrani, A. (2022, July). An ensemblebased machine learning for predicting fraud of credit card transactions. In Science and Information Conference (pp. 214-229). Cham: Springer International Publishing. [25] Mim, M. A., Majadi, N., & Mazumder, P. (2024). A soft voting ensemble learning approach for credit card fraud detection. Heliyon, 10(3). [26] Dolo, K. M. (2019). Differential evolution technique on weighted voting stacking ensemble method for credit card fraud detection (Doctoral dissertation, Doctoral dissertation). [27] Bonde, L., & Bichanga, A. K. (2025). Improving Credit Card Fraud Detection with Ensemble Deep Learning-Based Models: A Hybrid Approach Using SMOTE-ENN. Journal of Computing Theories and Applications, 2(3), 384. [28] Prabhakara, E., Kumarb, M. N., Ponnarb, K., Sureshb, A., & Jayandhiranb, R. (2019). Credit card fraud detection using boosted stacking. South Asian Journal of Engineering and Technology, 8(1), 149-153. [29] Buragadda, S., Phanindra, V. N., VenkateswarRao, S., & Goud, R. P. (2023, October). Hybrid Stacking Algorithm to Detect Fraudulent Transactions in Credit Card. In International Conference on Microelectronics, Electromagnetics and Telecommunication (pp. 269-281). Singapore: Springer Nature Singapore. [30] Ishak, N. A., Ng, K. H., Tong, G. K., Kalid, S. N., & Khor, K. C. (2022). Mitigating unbalanced and overlapped classes in credit card fraud data with enhanced stacking classifiers system. F1000Research, 11, 71.

Copyright

Copyright © 2025 Mrs. R. Jayalakshmi , Dr. R. G. Suresh Kumar, Thanushree T. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.