Credit Card Fraud Detection Using Machine Learning
Deep Prajapati, Ankit Tripathi, Jeel Mehta, Kirtan Jhaveri, Vishakha Kelkar
Abstract
From the day when payment systems emerged, there have been people willing to find novel ways to access someone’s finances illegally. This menacing hazards has grown in the current period, as the majority of transactions are now completed entirely online using credit card information. Frauds due to Credit Cards is a broad phrase that refers to any type of fraud involving a payment card, specifically a credit cards.The solitary purpose of such transgressions is usually to gain goods and services, or to make a huge payment to another account without the owner’s consent. According to the Nilson Report, By 2025, due to credit card fraud the United States has been projected to suffer losses up to 12.5 billion dollars. Using Machine learning algorithms to detect Credit card fraud is a process in which the data is investigated through various techniques to achieve the best possible outcomes in detecting and impeding fraudulent transactions. In order to evaluate different algorithms which accurately detect credit card fraud we have used techniques such as Random Forest, XGBoost, ANN (Artificial Neural Network). The results of these models can be used to effectively detect any credit card transaction happening whether a genuine one or fraudulent.