Predicting health insurance claim frauds using supervised machine learning technique

The healthcare sector is growing quickly, making it a complex system. At the same time, fraud in this sector is becoming a serious issue. Misuse of the medical insurance systems is one of the problems. The automated detection of healthcare scams uses data mining and machine learning approaches. In this work, an effort is made to provide a review of healthcare industry frauds and the methods for spotting them. A proposed solution to address issues in health insurance claims involves using a machine learning model that focuses on utilizing appropriate methods, identifying key data sources, and examining the properties of healthcare data. To gain insight into the patterns of the features, univariate and bivariate analysis are conducted on the data. Afterwards, proper data visualization is performed to determine which feature has the greatest impact. Based on the pre-processed data, a machine learning model is constructed and compared with four algorithms: Logistic Regression, Random Forest, Decision Tree Classifier, and Naive Bayes. The highest accuracy was achieved through the Decision Tree Classifier algorithm, with an accuracy of 97.03%.