Abstract :

Healthcare Fraud Detection Is A Critical Challenge Due To The Highly Imbalanced Nature Of Real-world Medical Datasets, Where Fraudulent Cases Are Significantly Fewer Than Legitimate Ones. Conventional Machine Learning Techniques Often Struggle To Accurately Identify Such Rare Events, As They Tend To Favor The Majority Class. Existing Resampling Methods, Including Random Oversampling (ROS), Random Undersampling (RUS), And SMOTE, Partially Address This Issue But Introduce New Limitations Such As Overfitting, Data Loss, And Noise Generation. To Overcome These Challenges, This Study Proposes An Enhanced Fraud Detection Framework Based On A Hybrid Resampling Technique, SMOTE-ENN, Which Combines Synthetic Data Generation With Noise Reduction. Additionally, Feature-driven Enhancement Using The “Provider Type” Attribute Is Incorporated To Improve The Representation Of Minority Classes In A More Meaningful Way. He Proposed System Is Evaluated Using Multiple Machine Learning Algorithms, Including Decision Tree, Random Forest, Support Vector Machine, K-Nearest Neighbors, Logistic Regression, And Naive Bayes. Performance Is Assessed Using Advanced Evaluation Metrics Such As Accuracy, Precision, Recall, F1-score, AUC-ROC, And AUPRC, Which Are More Suitable For Imbalanced Datasets. Experimental Results Demonstrate That The Hybrid SMOTE-ENN Approach Significantly Improves Model Performance. Among All Classifiers, The Decision Tree Model Achieves The Highest Accuracy Of 0.99, Indicating Its Effectiveness In Detecting Fraudulent Activities. The Study Highlights The Importance Of Combining Hybrid Resampling Techniques With Feature Engineering To Develop Accurate And Reliable Healthcare Fraud Detection Systems.

Published:

22-5-2026

Issue:

Vol. 26 No. 5 (2026)

Page Nos:

1477-1482

Section:

Articles

License:

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

How to Cite