Processed Data: Cleaned and prepared the dataset using techniques like PCA for dimensionality reduction.
Engineered Features: Developed and selected key features to enhance fraud detection model performance.
Evaluated Models: Tested and compared machine learning algorithms (Logistic Regression, Genetic Algorithm, Random Forest, XGBoost) and assessed performance with precision, recall, and F1-score metrics.
Implemented Real-time Prediction: Demonstrated fraud detection in real-time and addressed challenges of imbalanced datasets to improve model robustness.