Securing Credit: A Hybrid multi-dimensional model using ensemble machine learning classifier with data sampling to detect and prevent credit card fraud.

The current methods for detecting fraud have notable shortcomings, including issues with imbalanced datasets, incorrect detection of fraudulent activities, limited versatility across various contexts, and challenges in real-time data processing. This study introduces an ensemble machine learning model aimed at identifying fraud in credit card transactions. Additionally, it employs the Synthetic Minority Oversampling Technique (SMOTE) combined with Edited Nearest Neighbor (ENN) to tackle the challenge of imbalanced data. The results from our experiments indicate that this method outperforms existing approaches. Consequently, it lays a crucial foundation for ongoing research focused on creating more resilient and adaptable systems for fraud detection.