Credit Card Fraud Detection
Junior Data Scientist Project | Python, Scikit-learn
→
Summary
Built supervised and unsupervised ML models on a highly imbalanced dataset. Achieved high fraud detection recall by prioritizing minority class performance. Evaluated models using Recall, F1-score, ROC-AUC, and confusion matrices. Analyzed trade-offs between false positives and missed fraud cases. Demonstrated understanding of real-world ML limitations and evaluation metrics.