Healthcare Fraud Detection Advanced Analytics

Design a comprehensive anomaly detection system for identifying potential healthcare insurance fraud using advanced unsupervised machine learning techniques. Develop a multi-layered approach that integrates claim data, patient records, billing histories, and statistical pattern recognition. Create a robust scoring mechanism that can flag suspicious activities with minimal false-positive rates, incorporating ensemble learning methods and adaptive threshold algorithms.