Fake Digital Payment Detector

Authors:

Atharva Ravindra Chothe¹, Aman Barkat Shaikh², Suraj Manoj Dol³, Koustubh Chandrakant Mohite^4, Ashwini ashok mahind⁵

Assistant Professor, Computer Science and Engineering, Adarsh Institute of Technology and Research Centre, Vita, and Vita¹

1. Tech, Computer Science and Engineering, Adarsh Institute of Technology and Research Centre, Vita, and Vita²
2. Tech, Computer Science and Engineering, Adarsh Institute of Technology and Research Centre, Vita, and Vita³
3. Tech, Computer Science and Engineering, Adarsh Institute of Technology and Research Centre, Vita, and Vita⁴
4. Tech, Computer Science and Engineering, Adarsh Institute of Technology and Research Centre, Vita, and Vita⁵
5. Tech, Computer Science and Engineering, Adarsh Institute of Technology and Research Centre, Vita, and Vita⁶

Abstract: The rapid growth of digital payment systems has made online transactions fast and convenient, but it has also increased the risk of payment fraud. Traditional rule-based security mechanisms are often unable to detect complex and evolving fraudulent behaviours. This paper presents a Fake Digital Payment Detector, an intelligent system designed to identify suspicious and fraudulent transactions in real time.

The proposed system combines machine learning models with rule-based anomaly detection to analyse transaction patterns such as unusual payment amounts, frequent transactions, device or IP address mismatches, and abnormal spending behaviour. The system is developed using Python, Django, and Vue.js and is integrated with the Razor pay payment gateway to evaluate transaction risk scores before final payment approval. Multiple algorithms, including Logistic Regression, Random Forest, Gradient Boosting, Isolation Forest, and Local Outlier Factor, are used to build a hybrid fraud detection framework. Experimental results show improved detection accuracy with fewer false positives, ensuring secure and reliable digital payment processing.

Keywords - Fraud Detection, Machine Learning, Anomaly Detection, Real-Time Processing, Hybrid Detection Framework, Transaction Risk Scoring, Isolation Forest, Digital Payments