Detection of Fraudulent Transactions in the Banking Sector

The banking sector has a great influence on the modern world. The market is actively developing and growing, new banking products and services are being developed, new technologies are being introduced that increase interest of fraudsters in this industry. Every year, billions of dollars losses due to fraudulent bank card transactions are reported. The detection of bank card fraud is a typical example of a problem in data science, therefore, increasing attention is being paid to machine learning methods in fraudulent activities. Difficulties in the development of such models are associated with a highly unbalanced distribution of classes. The aim of this paper is to identify approaches based on training methods and data balancing methods that have created the most accurate models for determining fraudulent actions on bank cards.