UPI Fraud Detection System Using CNN Machine Learning Model

Unified Payments Interface (UPI) has revolutionized digital transactions in India offering convenience and real-time processing. However, this rapid adoption has also led to a surge in fraudulent activities, challenging the efficacy of traditional rule-based fraud detection methods. These conventional systems often struggle to adapt to evolving fraud patterns, necessitating more robust and adaptive solutions. In response to these challenges, researchers have explored machine learning techniques to detect fraudulent activities within UPI transactions. While existing strategies have shown promise, they are frequently validated on limited or synthetic datasets, which may not fully capture the complexities of real-world scenarios. To address these limitations, a comprehensive evaluation of prevalent machine learning classifiers—including Logistic Regression, K-Nearest Neighbors (KNN), Support Vector Machines (SVM), Naive Bayes, Decision Trees, Random Forests, and Convolutional Neural Networks (CNNs)—was conducted. Building upon this analysis, we propose the Convolutional Neural Networks (CNNs) framework, designed to enhance fraud detection capabilities in UPI transactions.