Machine Learning-Based Real-Time UPI Fraud Detection System
TL;DR
This project presents an ML-Based Real-Time UPI Fraud Detection System that uses machine learning algorithms to identify fraudulent transactions efficiently and shows that the Random Forest algorithm achieves the highest accuracy, making it the most effective model for fraud detection.
Abstract
ABSTRACT The rapid growth of Unified Payments Interface (UPI) transactions has increased the risk of online payment fraud. This project presents an ML-Based Real-Time UPI Fraud Detection System that uses machine learning algorithms to identify fraudulent transactions efficiently. The system preprocesses transaction data by handling missing values, encoding categorical features, and splitting the dataset for training and testing. Machine learning models such as Support Vector Machine (SVM), Random Forest, and AdaBoost are trained to classify transactions as genuine or fraudulent. The models are evaluated using accuracy, confusion matrix, and classification metrics. Experimental results show that the Random Forest algorithm achieves the highest accuracy, making it the most effective model for fraud detection. The proposed system enhances the security of digital payment platforms by enabling real-time fraud detection, reducing financial losses, and improving the reliability of UPI transactions. Keywords: UPI, Fraud Detection, Machine Learning, Random Forest, Real-Time Detection, Digital Payments.