Enhanced Deep Neural Network with SMOTE for Credit Card Fraud Detection

In the digital economy of today where businesses profit from electronic transactions, credit card fraud is a serious problem that causes millions of dollars of losses every year. To overcome this problem, our research proposes a novel deep learning-based solution to detect fraudulent transactions effectively. To mitigate this issue, the proposed model applies the Synthetic Minority Oversampling Technique (SMOTE) due to the common class imbalance issue seen in realistic transaction data. The model is trained using a publicly available data set containing anonymized transaction records with a deep neural network optimized with the Adam optimizer and with ReLU activation. We evaluate the performance with important classification metrics (accuracy, precision, recall, F1-score, and AUC-ROC curve). The results demonstrate that the improved deep learning algorithm works better than traditional machine learning techniques, offering a dependable and useful strategy for identifying financial system fraud.