Using a Bayesian Belief Network to Detect Healthcare Fraud

May 12, 2026

Healthcare fraud detection in pharmacy provider systems requires models that integrate with real business workflows while remaining robust to changing data and user requirements such as missing inputs or scaling needs. This study proposes a Bayesian Belief Network (BBN) framework as a robust and interpretable alternative to conventional anomaly and fraud detection approaches, aiming for comparable predictive performance. The work details model development, parameter learning, data preprocessing, and feature engineering, then compares results to baseline methods using quantitative evaluation. Interpretability is assessed through qualitative, auditor- and investigator-oriented common-sense analysis.

Share with

May 12, 2026