Islam, Md Shakil and Rahman, Nayem (2026) AI-Driven Detection of Medicaid and Medicare Billing Fraud to Reduce U.S. Healthcare Waste and Economic Loss. American Journal of Economics and Business Management, 9 (9). ISSN 2576-5973

Text AJEBM+15478+AI-Driven+Detection+of+119-142.pdf Download (1MB)

Abstract

A major problem in the U.S. healthcare system is healthcare billing fraud in Medicare and Medicaid programs that inflict billions of dollars in financial losses and add to healthcare waste every year. Duplicate billing, phantom claims, upcoding, unnecessary medical procedures, and patient information misuse all pose detrimental consequences on the efficiency of the healthcare system and resources allocation. Most fraud detection techniques are still manual and rule-based, and are inefficient, time consuming, and lack the ability to detect complex fraud patterns in large scale healthcare data. So, there is a rising demand for a smart and automated fraud detection system that can increase the accuracy of fraud detection and minimize economic losses. This research presents an Artificial Intelligence (AI) based approach to identify Medicaid and Medicare billing fraud through the use of sophisticated machine learning (ML) methods. This study uses the Healthcare Provider Fraud Detection Analysis data set, which comprises records of inpatient and outpatient claims and records for Medicare beneficiaries in the United States. To enhance the data quality and model performance, data preprocessing techniques such as data cleaning, normalization, feature engineering, and addressing class imbalance are applied. Multiple machine learning algorithms like Random Forest, Logistic Regression, Decision Tree, Support Vector Machine and XGBoost are used and evaluated for performance, to select the best model for healthcare fraud detection. The proposed system seeks to automatically detect suspicious billing behaviors, as well as to accurately categorize fraudulent claims with a minimum number of false positive predictions. Performance of the model is assessed by the accuracy, precision, recall, F1 score, ROC-AUC and confusion matrix analysis. The potential outcomes of this research are an effective AI-powered healthcare fraud detection system that can help curb healthcare waste, lower financial loss, enhance claim verification, and boost financial security across the U.S. Medicare and Medicaid programs. The study also showcases how AI can be leveraged to reshape healthcare fraud detection and enhance the effectiveness of healthcare programs in public healthcare systems.

Item Type:	Article
Uncontrolled Keywords:	Healthcare Fraud Detection, A Medicare and Medicaid Fraud resource for healthcare professionals, Artificial Intelligence, Machine Learning, Predictive Analytics and Healthcare Waste Reduction
Subjects:	H Social Sciences
Depositing User:	admin eprints
Date Deposited:	07 Sep 2026 08:16
Last Modified:	07 Sep 2026 08:16
URI:	http://eprints.umsida.ac.id/id/eprint/17062

Actions (login required)

View Item