Integrating Metaheuristics in Computer Vision for Real-World Optimization Problems

4Comparative Evaluation of Machine Learning Algorithms for Bank Fraud Detection

Kiran Jot Singh^1,2, Divneet Singh Kapoor^1,2*, Kunal Ranjan Singh³, Chirag Kalucha³, Gatik Alagh³, Khushal Thakur^1,2 and Anshul Sharma^1,2

¹Kalpana Chawla Centre for Research in Space Science & Technology, Chandigarh University, Mohali, Punjab, India

²Electronics and Communication Engineering Department, Chandigarh University, Mohali, Punjab, India

³Computer Science and Engineering Department, Chandigarh University, Mohali, Punjab, India

Abstract

Digital fraud is increasing globally with the rapid surge in digital transactions. This makes every transaction through digital mediums prone to fraud, creating a significant problem on a global scale. Online payment frauds are estimated to cause the world economy to lose about 48 billion US dollars of e-commerce by 2023, and approximately 1.7 million people worldwide are affected by malicious digital banking frauds annually. The previous approach of using suspicious activity reports (SARs) to detect banking frauds has proven to be less efficient. To address this issue, a proposed solution is based on a classification model trained on a dataset. This model enables the device to differentiate between a fraud and a valid transaction by analyzing their domains and training knowledge. The selection of the classification model was made after a comparative evaluation of several models, with the most accurate and precise one used for final testing after training ...