Big data techniques can be used to detect Medicare fraud. The document discusses building a machine learning model to predict fraud using Medicare prescription drug, payment, and exclusion data. Several fraud patterns and government efforts are outlined. Exploratory data analysis was performed on the datasets and features were engineered before various classification models were trained. The best performing model was random forest, achieving an AUC of 72%. Future work could involve model retraining without stopping predictions using real-time streaming data and Kafka.