Health Care Provider Fraudulent Detection Using Machine Learning

Healthcare fraud is a serious problem that affects the financial health and trust in healthcare systems around the world. This research paper focuses on using machine learning to detect fraudulent activities by healthcare providers. We analyze large amounts of data from Medicare claims to find unusual patterns that may indicate fraud. By using different machine learning methods, such as decision trees and random forests, we create a model that can accurately separate legitimate claims from fraudulent ones. To tackle the challenge of imbalanced data, we apply techniques like oversampling, which helps improve our model's performance. Our results show that this machine learning approach significantly enhances the accuracy and reliability of fraud detection compared to traditional methods. Additionally, our findings provide valuable insights for healthcare administrators and policymakers, helping them take action against fraud more effectively. By incorporating these advanced techniques into existing systems, we aim to support efforts to protect healthcare resources and improve patient care. This research not only adds to the understanding of fraud detection in healthcare but also offers practical solutions to fight against it effectively. Keywords: Provider Review, Insurance Claim Detection, Supervised Machine Learning, Support Vector Machine, Random Forest.