Variables Creation in Fraud Detection-Based on New York Property Data

Abstract

With the rapid development of technology, fraud has become an extremely serious problem in the US society, therefore makes fraud detection an urgent need to improve the situation as much as possible. Fraud detection refers to the process of finding anomalies in a huge bunch of data by building fraud algorithms and models that could predict the possible behaviour of the real world situation. This paper will mainly focus on the 'Variables Creation' process in fraud detection with the New York property data and discuss how to analyze a fraud problem and build variables based on New York property data. The whole process of fraud detection will render practical help to solve the property fraud problem.