Angelina Sidorov Data Analytics track

Healthcare Provider Fraud Risk Analysis Project

Analyzed 40,000 anonymized Medicare claim records across 5,000 healthcare providers, performing data cleaning and feature engineering. Trained and evaluated multiple machine learning classification models in Python and applied anomaly detection to identify provider patterns associated with potential fraud. Developed a multi-stage fraud analysis pipeline combining risk scoring, geographic insights and provider-level explanations to improve interpretability of model outputs.