Random forest for credit card fraud detection

TLDR

Credit card fraud frequently causes large financial losses, with criminals exploiting technologies like Trojans or phishing to steal card information. The study aims to evaluate two random forest approaches for detecting credit card fraud by training on historical transaction data. The authors train two random forest models on historical transaction data from a Chinese e‑commerce company, extracting normal and fraud behavior features, and compare the models’ performance for fraud detection. The comparison revealed differences in detection performance between the two random forest variants.

Abstract

Credit card fraud events take place frequently and then result in huge financial losses. Criminals can use some technologies such as Trojan or Phishing to steal the information of other people's credit cards. Therefore, an effictive fraud detection method is important since it can identify a fraud in time when a criminal uses a stolen card to consume. One method is to make full use of the historical transaction data including normal transactions and fraud ones to obtain normal/fraud behavior features based on machine learning techniques, and then utilize these features to check if a transaction is fraud or not. In this paper, two kinds of random forests are used to train the behavior features of normal and abnormal transactions. We make a comparison of the two random forests which are different in their base classifiers, and analyze their performance on credit fraud detection. The data used in our experiments come from an e-commerce company in China.