by The estimated annual fraud exceeds $100 billion and is likely even higher. Traditional methods of fraud detection involve a limited number of auditors manually inspecting thousands of claims, which can only identify specific patterns of suspicious behavior. However, there are not enough investigators to keep up with the various Medicare fraud schemes.
To address this challenge, researchers at the College of Engineering and Computer Science at Florida Atlantic University have developed a new AI technique that can pinpoint fraudulent activity in big Medicare data. This technique utilizes big data from patient records and provider payments to create effective machine learning models for fraud detection.
The researchers tested their technique on two imbalanced big Medicare datasets: Part B and Part D. Part B covers medical services like outpatient care, while Part D relates to prescription drug benefits. The datasets were labeled with the List of Excluded Individuals and Entities (LEIE) provided by the United States Office of the Inspector General.
The research focused on the influence of two techniques: Random Undersampling (RUS) and supervised feature selection. RUS is a data sampling technique that randomly removes samples from the majority class until a balance between the minority and majority classes is achieved. Supervised feature selection involves selecting the most relevant features to improve classification accuracy.
The experimental design explored various scenarios, including the use of each technique individually and in combination. The researchers found that the combination of RUS and supervised feature selection outperformed models that utilized all available features and data. The technique of performing feature selection followed by RUS yielded the best performance.
The results demonstrated that intelligent data reduction techniques can significantly improve the classification of imbalanced big Medicare data. By reducing the number of features, the models become more explainable and perform better than models using all features.
This novel technique has the potential to revolutionize Medicare fraud detection by conserving substantial resources for the Medicare system. It can automate the detection process and identify fraudulent activities that would otherwise go undetected. By utilizing AI technology, the Medicare system can stay one step ahead of criminals, safeguarding funds and ensuring that healthcare providers operate ethically.
The findings of this study have been published in the Journal of Big Data and offer a promising solution to the persistent problem of Medicare fraud. With further development and implementation, this AI technique could make a significant impact in preventing fraudulent insurance claims and protecting the integrity of the Medicare system.
Note:
1. Source: Coherent Market Insights, Public sources, Desk research
2. We have leveraged AI tools to mine information and compile it
