<ul><li>Healthcare fraud costs billions of dollars annually. Fraudulent providers overcharge insurance, submit inflated claims, or game the reimbursement system.</li><li>In this project, a supervised machine learning pipeline was built to identify high-risk providers based on claim-level data, with a focus on accuracy, recall, explainability, and handling class imbalance.</li><li>Key features were engineered to reflect billing patterns, visit frequency, and reimbursement behavior, aiming to distinguish honest providers from those likely gaming the system.</li><li>XGBoost provided the best tradeoff between precision and recall, making it the recommended model for deployment in fraud detection in healthcare.</li></ul>

Machine Learning for Fraud Detection in Healthcare

Discover more