Credit Card Fraud Detection Using Optimized Ensemble Learning Models

Abstract

Credit card fraud poses a significant threat to financial institutions and their clientele worldwide. Many current fraud detection methods lag behind the evolving strategies of fraudsters, resulting in escalating financial risks. This study outlines a robust method for detecting credit card fraud, employing optimal ensemble learning techniques. The proposed approach integrates ensemble learning principles into both stacking and boosting methodologies, utilizing model parameters to bolster the accuracy and resilience of fraud detection. The research investigates various ensemble learning algorithms and evaluates their effectiveness in identifying fraudulent transactions using real transactional data. The experimental results demonstrate the ability of optimized ensemble learning models to detect fraudulent activities, underscoring their potential to enhance security measures against credit card fraud