Loading…
Welcome to the ARIA 2021 Annual Meeting Agenda App. You are seeing this in simple view. Click on any session to see Presenters, Discussants and a Session Description. Click on Presenters to find a single presenter and their sessions. If you find an error, please email gphillips@aria.org. Registrants have been added and will continue to be updated until 7/29/21 when registration closes. Zoom links will be shared directly with all participants two days prior to the conference. Please make notifications@sched.com a "safe sender" in your email system. We will be sending messages throughout the conference through this medium. 
Back To Schedule
Tuesday, August 3 • 2:30pm - 4:00pm
5D2 Physician Fraud Detection: The Machine Learning Approach

Sign up or log in to save this to your schedule, view media, leave feedback and see who's attending!

Feedback form is now closed.
Gene Lai, University of North Carolina at Charlotte; Licheng Jin, Southwestern University of Finance and Economics; Chia-Ling Ho, Tamkang University

Using the supervised machine learning method, we try to detect fraudulent physicians. We apply the supervised neural network to predict the likelihood of a physician to be fraudulent. We choose the Oversampling, the Synthetic Minority Oversampling Technique (SMOTE), and the hybrid methods to deal with the imbalanced data issue. Our classifier provides AUROC (Area Under the Receiver Operating Characteristic Curve) scores approximately 0.80 for these three methods, indicating that the methods can largely separate the fraudulent physicians from the legitimate ones. Using the threshold 0.5, our methods generate the GMean 0.7015, 0.7071, and 0.7304 for the Oversampling, SMOTE, and Hybrid methods, respectively. The Recall equals 0.6829, 0.6585, and 0.6098 for these three methods accordingly. Compared with the traditional logistic regression, our method is more appropriate for physician fraud detection since the logistic regression underfits the data. This paper has important implications for insurers: speeding up the claim review process, narrowing the fraud investigation range, excluding suspicious fraudulent physicians as external reviewers, and accurately identifying fraudulent physicians. Moreover, our results also have policy implications for policyholders, hospitals, and the health care industry because fraudulent claims cost all the parties involved.

Discussant
Presenters
GL

Gene Lai

University of North Carolina-Charlotte



Tuesday August 3, 2021 2:30pm - 4:00pm EDT