Applying clustering and AHP methods for evaluating suspect healthcare claims

This paper seeks to present a model for the analysis of suspicious claims data from healthcare providers with the use of different clustering algorithms, and the application of the AHP multicriteria method for prioritizing the identified suspect entities for subsequent auditing. We begin with a brief overview of related works that have covered the application of the aforementioned techniques for investigating suspicious entities in the context of internal auditing and healthcare. After presenting the steps for the construction of our own model, we discuss our results. We determine that the application of clustering algorithms to our initial variables resulted in the automatic detection of almost all entities initially classified as suspect. Our AHP model then provided us with rational criteria for effectively and objectively ranking these entities for further investigation