Fraud Detection for Credit Card Transactions Using Random Forest Algorithm

Abstract

In these days, credit card fraud detection is a major concern in the society. The use of credit cards in e-commerce sites and various banking sites has been increased rapidly in recent times. As modernization will have both positive and negative impacts, the use of credit cards in online transactions has made it simple; likewise, it also led to the increase of the number of fraud transactions. As part of the activities happening, it is always advised for the e-commerce sites and the banks to have automatic fraud detection system. Credit card fraud might result in huge financial losses. While look for the solutions for credit card frauds that are happening, machine learning techniques provide favorable solutions. The proposed system uses a random forest application in solving the problem and to attain more accuracy when compared to the other algorithms used till now. All the basic classifiers have the same weight but random forest algorithm has relatively high and others have relatively low weights because of the randomization of bootstrap sampling of a making decision and selection of attributes cannot guarantee that all of them have the same stability in decision making. KeywordsDecision treeFraud detectionRandom forest