The effects of varying class distribution on learner behavior for medicare fraud detection with imbalanced big data

Abstract

Healthcare in the United States is a critical aspect of most people's lives, particularly for the aging demographic. This rising elderly population continues to demand more cost-effective healthcare programs. Medicare is a vital program serving the needs of the elderly in the United States. The growing number of Medicare beneficiaries, along with the enormous volume of money in the healthcare industry, increases the appeal for, and risk of, fraud. In this paper, we focus on the detection of Medicare Part B provider fraud which involves fraudulent activities, such as patient abuse or neglect and billing for services not rendered, perpetrated by providers and other entities who have been excluded from participating in Federal healthcare programs. We discuss Part B data processing and describe a unique process for mapping fraud labels with known fraudulent providers. The labeled big dataset is highly imbalanced with a very limited number of fraud instances. In order to combat this class imbalance, we generate seven class distributions and assess the behavior and fraud detection performance of six different machine learning methods. Our results show that RF100 using a 90:10 class distribution is the best learner with a 0.87302 AUC. Moreover, learner behavior with the 50:50 balanced class distribution is similar to more imbalanced distributions which keep more of the original data. Based on the performance and significance testing results, we posit that retaining more of the majority class information leads to better Medicare Part B fraud detection performance over the balanced datasets across the majority of learners.

Survey on deep learning with class imbalance

“…In this example, the majority group of healthy patients is referred to as the negative class. Learning from these imbalanced data sets can be very difficult, especially when working with big data [8,9], and non-standard machine learning methods are often required to achieve desirable results. A thorough understanding of the class imbalance problem and the methods available for addressing it is indispensible, as such skewed data exists in many real-world applications.…”

Ensemble-AMPPred: Robust AMP Prediction and Recognition Using the Ensemble Learning Method with a New Hybrid Feature for Differentiating AMPs

“…Therefore, developing well-balanced training data is an important step. There are two categories of methods that handle this class imbalance problem: Data-Level methods (e.g., data sampling), and Algorithm-Level methods (e.g., cost-sensitive and hybrid/ensemble) [ 56 , 57 , 58 ].…”

What Should Health Care Organizations Do to Reduce Billing Fraud and Abuse?

“…3 Using real-time data collection, the Office of the Inspector General can compare patient volume for similar professional claims to identify abnormally high reimbursement submissions, unnatural practice growth patterns, or unusually high numbers of procedures based on specialty and practice size or to flag suspect patient visits patterns (such as an excessive number of patients during a 24-hour window.) 22,23 This artificial intelligence-based system for identifying potential program integrity anomalies is relatively new. But CMS is also directed to cases by whistleblowers, who are incentivized to report fraud under the False Claims Act and Stark Law (ie, prohibition on self-referral), which entitle them to receive a percentage of any government recoveries.…”