Investigating and Identifying Fraudulent Behaviors of Medical Claims Data Using Machine Learning Algorithms

Abstract

Healthcare is essential in pandemic times, but it is crucial for the well-being of daily life. Many countries allocate substantial funds towards providing high-quality healthcare services. As healthcare expenses escalate, policymakers and funders are increasingly focused on investigating the underlying factors driving the high costs of medical resources. A comprehensive analysis carried the required expenses towards identification, valuation, and measurement of resources utilized for the diagnosis process. The objective of the chapter is to provide how the data analysis is carried which helps to identify fraudulent behaviors. The generated model assists health management organizations in identifying suspicious behaviors toward claims. Healthcare fraud is a severe threat to global health results, and could lead to misuse, scarce resources, and negative impacts on healthcare access, infrastructures, and social determinants of health. Healthcare fraud is associated with increased healthcare costs in most of the leading countries. The proposed research work provides an estimation mechanism for utilizing health resources and their impacts on healthcare costs. This chapter proposes strategic ways of handling healthcare data to prevent future healthcare fraud, decrease healthcare expenditure, and adequately use resources to benefit the population. This chapter works on three primary datasets and a synthetic dataset aggregated from the primary datasets. The data preprocessing is carried out at different levels of the model, which truly enhances the data quality. The model is constructed at three levels; the first level analyzes datasets in which it extracts the primary features and provides constructive decisions and outcomes on the processing of data. Regressive analysis of the hierarchical grouping mechanism helps to know the detailed features that could affect healthcare and prevent resource misuse.