A Comparative Study of XGBoost and Logistic Regression for Fraud Detection in Online Transactions

In today's digital era, electronic transactions play a vital role in everyday business, banking, and personal finance. However, this rise in convenience has also led to a significant increase in online fraud. Cybercriminals use techniques such as phishing, identity theft, and automated bots to exploit vulnerabilities in payment systems, leading to financial losses and eroding consumer trust. These fraudulent activities not only impact individuals but also disrupt the operations of banks and e-commerce platforms around the world. Traditional rule-based detection systems often struggle to keep up with the complexity and scale of modern transactions. To address this, our study explores the use of machine learning techniques-specifically XGBoost and Logistic Regression-for detecting fraudulent transactions. These models are trained on real-world transaction data and evaluated for their effectiveness in identifying suspicious activity. The results demonstrate the potential of machine learning to enhance the accuracy and reliability of fraud detection systems.