Discovery of fraud in Medical Insurance
Abstract
The term insurance fraud refers to the commission of any act with the intent to obtain an outcome that is favorable, but fraudulent during an insurance claim. Including single prescription medicines is extremely high, card repeatedly within a certain amount of time for medicine, etc. This paper is based on methods of hierarchical cluster analysis and generalized squared distance discriminate method to record medical coverage of transaction data at outliers for finding out the corresponding abnormal record which indicates potential fraud.
Keywords
Download Options
Introduction
Insurance fraud is very serious and widespread violations in the insurance industry. This paper studies the methods of data preprocessing, visualization, hierarchical cluster analysis and generalized squared distance discrimination.
Conclusion
In this paper, the samples are divided into one class by cluster analysis, after that, the most similar two samples are clustered into a small class, and then merge with the most similar small classes. The Euclidean distance we used is not affected by the dimension. The Euclidean distance between two points has nothing to do with the measuring unit of the original data. The distance is the same between the two points which is calculated by the standardized data and the center data. This discovery has a certain reference value for the discovery and research of medical insurance fraud.