Healthcare Insurance Fraud Detection Through Data Mining Techniques

Abstract

Healthcare programs and insurance initiatives play a crucial role in ensuring that people have access to medical care. Countries as well as corporate companies around the world provide healthcare facilities to their citizen and employees for the balance and healthy life. However, despite the benefits of these programs, healthcare insurance fraud continues to be a significant challenge in the industry. Reports says, amount worth more than $760 Billion wasted every years in terms of insurance fraud in United States. In this study, we present a model that utilizes five unsupervised learning techniques to detect healthcare insurance fraud. We used the Centers for Medicare and Medicaid Services (CMS) 2008-2010 DE-SynPUF dataset for our analysis. Our model began by implementing the Apriori Association Rule Mining Technique to extract frequent rules from the dataset. We then passed the extracted rules to fraudulent classifiers, such as IF, CBLOF, ECOD, and OCSVM, to identify fraudulent activity. However, while our model demonstrated potential, further research and testing are necessary to improve its effectiveness and accuracy. The healthcare industry generates vast amounts of data, and a more extensive analysis of multiple healthcare insurance datasets could improve our model’s performance. Machine learning solutions offer the possibility of significantly reducing fraudulent activity in the healthcare industry, which could result in improved patient care and reduced healthcare costs.