Improving Credit Card Fraud Detection Using Ensemble Machine Learning Techniques

<title>Abstract</title> <p>The rapid growth of online transactions has significantly increased the risk of credit card fraud, making accurate and real-time fraud detection systems essential for financial platforms. This paper proposes an enhanced fraud detection framework that combines Generative Adversarial Networks (GANs) for data balancing with ensemble machine learning techniques to improve detection performance. The study is conducted on a real-world dataset containing 76,900 transactions, where fraudulent operations represent only 2.5\% of the total data, leading to a severe class imbalance problem. To address this issue, a GAN-based approach is employed to generate synthetic samples and create a balanced training dataset. Several machine learning models, including XGBoost, Random Forest, Logistic Regression, SVM, KNN, and Naïve Bayes, are evaluated using performance metrics such as Accuracy, Precision, Recall, F1-score, and AUC-ROC, along with execution time to assess real-time feasibility. Experimental results demonstrate that ensemble models, particularly XGBoost, achieve superior performance, significantly improving fraud detection recall while maintaining high precision and low latency. The findings confirm that integrating synthetic data generation with ensemble learning provides a robust, scalable, and practical solution for real-world credit card fraud detection systems.</p>