Use of Big Data Technologies for Credit Card Fraud Prediction

Abstract

Globalization has led to an increased usage of credit cards as the mode of payment for various kinds of transactions, both online as well as offline. Just as a coin has two sides, Globalization also has another side to it – increased affinity of the Banking industry to fall prey to fraud. The main objective of this paper is to study the best methodology to predict credit card fraud accurately. By utilizing big data technologies and Machine learning algorithms (Logistic Regression, Decision Tree, Random Forest, etc.), fraudulent transactions in credit card operations can be predicted in advance to mitigate further risk. This paper outlines the best pre-processing practices for accurate results of credit card fraud prediction. The secondary mode of Qualitative research is chosen for this research paper. This study would benefit the banking industry, credit card companies, and academicians by helping them understand how algorithms can be refined to detect and mitigate fraudulent credit card practices.