Deep Learning Techniques for Financial Fraud Detection

Financial fraud is a non-negligible issue in the finance industry and cause considerable cost for government, banks, and ordinary card holders. Classical methods of financial fraud detection demand massive manual work, even if time and economic cost are expensive the accuracy of detection is dissatisfied. However, with the development of Machine Learning technique and cloud computing in recent years provide new methods to detect financial fraud. The most common technique is machine learning there are many research about using machine learning model to detect financial frauds. Use deep leaning to detect financial will significantly reduce the cost and increase the detect accuracy. In the past decade, researchers used different machine learning models and apply both supervised and unsupervised learning to solve problems occur in fraud detection task, but still lack of comprehensive classification and conclusion about financial fraud detection. This review will provide comprehensive analysis on existing methods of fraud detection literature base on the core aspects and general conclusion about the development of the detection methods.