Identifying Credit Card Frauds Through Data Driven Machine Learning Models: A Review

Identifying Credit Card Frauds Through Data Driven Machine Learning Models: A Review

Naman Bhawre1, Prof. Sanmati Jain2

Research Scholar1, HOD (CSE & IT)2

VITM, Indore, India1,2

Abstract—Credit card frauds has become extremely prevalent with increase in online transactions. The issue is especially critical in mobile devices due to the less sophisticated filtering mechanisms in their built-in operating systems. Recent improvements in electronic commerce and communication technology have markedly enhanced the utilization of credit cards for both online and conventional transactions. Nonetheless, there has been a consistent increase in fraudulent credit card transactions, resulting in substantial annual losses for financial institutions. The creation of efficient fraud detection algorithms is essential for reducing these losses; nevertheless, it is difficult due to the significant imbalance in most credit card datasets. Conventional rule-based systems frequently lack the capability to address the complex and adaptive methods employed by fraudsters. Machine learning (ML) offers more dynamic, scalable, and efficient techniques for identifying fraudulent activities. This study provides a thorough examination of credit card datasets, the uneven characteristics of these datasets, and the current baseline methodologies in the field..

Keywords— Credit Card Fraud Detection, Daat Driven Models, Machine Learning, ,  Imbalanced  Datasets, Classification Accuracy