Effective fraud detection in healthcare domain using popular classification modeling techniques

Abstract

Fraud is any activity with malicious intentions resulting in personal gain. In the Present Day scenario, every sector is polluted by such fraudulent activities to fetch unauthorized benefits. In HealthCare, an increase in fraudulent insurance claims has been observed over the years which may constitute around 3-5% of the total cost. Increasing healthcare costs along with the hike in fraud cases have made it difficult for people to approach these services when required. To avoid such situations, we must understand and identify such illegal acts and prepare our systems to combat such cases. Thus, there is a need to have a powerful mechanism to detect and avoid fraudulent activities. Many Data mining approaches are applied to identify, analyze and categorized fraud claims from the genuine ones. In this paper, various frauds existing in the Health Care sector have been discussed along with analyzing the effect of frauds in the health care domain with existing data mining models. Furthermost, a comparative analysis is performed on two existing approaches to extract relevant patterns related to fraudulent claims.