Credit Card Fraud Detection Using Isolation Forest

Abstract

Nowadays credit card use has become extremely common. Generally, credit card fraud activity can happen both online and offline. Nowadays most people use online transaction due to which increasing in online transactions by using different payment methods, such as credit/debit card PhonPe, Gpay, Paytm, etc., fraudulent activities have also increased. Credit card fraud stands as a major problem for the world financial institute. According to an RTI report 2480 cases of fraud in 18 public sectors involving Rs. 31, 898, 63. According to RBI in 2017-2018 total 911 credit card fraud amounting to 65.6 crore. The acceptance and rejection of a transaction process happens within a micro or millisecond. Therefore, the detection of a fraud transaction must be extremely quick and effective. There are more than a million transactions which occur daily, and it is difficult to monitor each transaction individually. Thus, an effective fraud detection system is used to differentiate genuine and a fraud transaction. Our project plan to illustrate the design of a data set using machine learning with Credit Card Fraud Detection. The Credit Card Fraud Detection Problem includes design past credit card transactions with the data of ones that turned out to be fraud. By using this model, we recognize whether a new transaction is fraudulent or not.