Published Paper Details:

As the usage of credit cards for online transactions has increased, so has the potential for credit card misuse and fraud, which can result in significant financial losses for both cardholders and financial institutions. This research study aims to detect credit card fraud, taking into account the challenges posed by publicly available data, imbalanced datasets, evolving fraud tactics, and high rates of false alarms. The literature review highlights various machine learning-based approaches for credit card fraud detection, including Extreme Learning Method, Decision Tree, Random Forest, Support Vector Machine, Logistic Regression, and XG Boost. An empirical analysis is conducted using the European card benchmark dataset to assess the effectiveness of these approaches. The machine learning algorithm is applied to the dataset, resulting in improved fraud detection accuracy. Further experiments are conducted by varying the number of hidden layers, epochs, and using the latest models, resulting in improved accuracy, f1-score, precision, and AUC curves with optimized values of 99.9%, 85.71%, 93%, and 98%, respectively. The proposed model outperforms advanced machine learning models for credit card fraud detection, even when applied to imbalanced datasets. The proposed approaches can be implemented effectively for real-world credit card fraud detection, including balancing the data to improve detection accuracy.