Advanced Fraud Detection: Leveraging K-SMOTEENN and Stacking Ensemble to Tackle Data Imbalance and Extract Insights

ABSTRACT This design presents a robust system for detecting credit card fraud using a mongrel model that combines Random Forest with K- means SMOTEENN, a fashion specifically designed to attack class imbalance. The approach intelligently creates synthetic nonage samples and removes noise to insure a cleaner, more balanced dataset. By integrating Random Forest — a model known for its rigidity to complex data and resolvable AI( LIME), the system becomes both accurate and interpretable. This combination not only enhances vaticination trustability but also offers translucency, making it a strong result for fraud discovery in finance. KEYWORDS Fraud Detection, Random Forest, K- SMOTEENN, Data Imbalance, Stacking Ensemble, resolvable AI, LIME