Applications of Bayesian Methods in Detection of Healthcare Frauds

The term fraud refers to an intentional deception or misrepresentation made by a person or an entity, with the knowledge that the deception could result in some kinds of unauthorized benefits to that person or entity. Fraud detection, being part of the overall fraud control, should be automated as much as possible to reduce the manual steps of a screening/checking process. In the health care systems, fraud has led to significant additional expenses. Development of a cost-effective health care system requires effective ways to detect fraud. It is impossible to be certain about the legitimacy of and intention behind an application or transaction. Given the reality, the best cost effective option is to infer potential fraud from the available data using mathematical models and suitable algorithms. Among these, in recent years coclustering has emerged as a powerful data mining tool for analysis of dyadic data connecting two entities. An important data mining task pertinent to dyadic data is to obtain a clustering of each entity. In this paper application of Bayesian ideas in healthcare fraud detection will be presented. The emphasis will be on the use of Bayesian co-clustering and link analysis methodologies to identify potentially fraudulent providers and beneficiaries who have unusual group memberships. Detection of such unusual memberships will be helpful to decision makers for auditing of providers and beneficiaries.