FRAUD DETECTION IN BANKING TRANSACTIONS USING MACHINE LEARNING

Abstract

With the exponential rise in digital banking and online transactions, fraud in the financial sector has become a major concern. Traditional rule-based fraud detection methods are often inadequate in identifying complex, evolving fraudulent patterns. This project aims to implement a machine learning-based fraud detection system that can accurately and efficiently detect fraudulent banking transactions. By training on historical transaction data, the system learns behavioral patterns and flags anomalies that deviate from normal activities. The model incorporates supervised algorithms like Random Forest, Logistic Regression, and XGBoost, which are trained on features such as transaction amount, location, time, and customer behavior. The proposed system helps financial institutions reduce fraud losses and improve customer trust by providing real-time, intelligent, and scalable fraud detection capabilities