Detection of Credit Card Fraudulence with Machine Learning Techniques and Meta-Heuristic Algorithms

The influence of the COVID-19 epidemic and the latest developments in e-payment technology have resulted in a notable increase in the volume of daily credit card purchases and digital purchases. There has additionally been a rise in credit card fraud, which is significantly affecting banks, businesses, and, lastly, retailers and sellers. An effective credit risk administration method helps lenders and banks distinguish between creditworthy and non-creditworthy customers. Credit is essential for a nation's economic survival and growth, as evidenced by historical statistics. Therefore, the implementation and establishment of appropriate procedures that can ensure the security of web-based card payments are urgently needed. Reducing financial declines and guaranteeing safe payments depend on effective fraud detection. To tackle the problem of detecting credit card fraud, the study described in the article suggests a metaheuristic algorithm using machine learning (ML) techniques. The proposed approach was evaluated using a database generated from European consumers. The findings of the study demonstrate that the proposed approach outperforms the other ML methods and has a significant recognition level for correctly classifying fraud. The outcome demonstrated a significant increase in system performance, with a 98 % categorization accuracy and a <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{9 1 \%}$</tex> feature dimension reduction.

Перевод пока недоступен