Credit Cards Frauds and Cybersecurity Threats Machine Learning Detection Algorithms as Countermeasures

Credit and Debit cards have become the choice mode of payment online as a result of the proliferation of electronic transactions and advancement in Information and Communication Technology ICT . Because of the increased use of credit cards for payment online, the number of fraud cases associated with it has also increased scammers and fraudsters are stealing credit card information of victims online and thereby stealing their monies. There is the need therefore to stop or abate these frauds using very powerful fraud detection system that detects patterns of credit card frauds in order to prevent it from occurring. In this paper we x rayed the concept of credit card frauds and how they are carried out by fraudsters. Python 3.7.6 programming language, Jupyter Notebook 6.0.3 and Anaconda Navigator 1.9.12 were used as experimental test bed. Also, we implemented two different supervised machine learning algorithms on an imbalanced dataset such as Decision Tree and Random forest techniques. A comparative analysis of the credit card detection capabilities of these machine learning algorithms were carried out to ascertain the best detection algorithm using different performance evaluation metrics such as accuracy, precision, recall, f1 score, confusion matrix. Experimental results showed that Random Forest outperformed Decision Tree algorithm slightly in performance metrics used for performance evaluation. Obodoeze Fidelis C. | Oliver Ifeoma Catherine | Onyemachi George Olisamaka | Udeh Ifeanyi Frank Gideon | Obiokafor, Ifeyinwa Nkemdilim "Credit Cards Frauds and Cybersecurity Threats: Machine Learning Detection Algorithms as Countermeasures" Published in International Journal of Trend in Scientific Research and Development (ijtsrd), ISSN: 2456-6470, Volume-6 | Issue-7 , December 2022, URL: https://www.ijtsrd.com/papers/ijtsrd52440.pdf Paper URL: https://www.ijtsrd.com/computer-science/computer-security/52440/credit-cards-frauds-and-cybersecurity-threats-machine-learning-detection-algorithms-as-countermeasures/obodoeze-fidelis-c