Advanced Medical Billing Fraud Detection System

Create a comprehensive PostgreSQL framework for detecting potential medical billing fraud using advanced statistical and pattern recognition techniques. Design a system that analyzes billing records across multiple dimensions, implements machine learning-compatible aggregation techniques, and generates real-time fraud risk scores. Develop recursive queries to track billing patterns, create complex statistical models, and generate detailed audit trails for compliance reporting.