A Decision Tree and Logistic Regression Algorithm-Based Model for Predicting Crimes Committed by Health Insurance Fraudsters

Abstract

As the world's medical systems have matured over time, there are still unscrupulous medical institutions and participants in society who use illegal means to defraud medical insurance funds. However, the complexity of the people committing fraud, the difficulty of accounting for the actual amount of fraud in the healthcare system, and the variety of frauds make it impossible for the public security authorities to effectively combat them. In this context, this project aims to use pandas and Apriori association rule algorithms for data pre-processing and feature mining, to extract fields with high correlation, to create models related to health insurance fraud clues, and to filter out potential fraud suspects through the analysis of health care consultation information, combined with public security perception data, and to combat them effectively. This project applies advanced information technology to police fraudulent medical insurance fund clues mining, which can promote the advancement of policing in healthcare while facilitating the integration of advanced information technology with traditional fields, while safeguarding the people's basic right to health and survival.