Publication Date: 2024/03/26
Abstract: Credit card fraud is an easy target. E- commerce and many other online sites collect money online, increasing the risk of online fraud. As fraud increases, researchers have begun using different learning techniques to detect and analyze fraud in online businesses. The main goal of this article is to design and develop a new streaming data transfer fraud method that aims to identify customer context and extract behavior from the business. Card holders are divided into different groups according to transaction fees. Sliding windows are then used to combine transactions from different cardholder groups, allowing the behavior of each group to be separated. The different groups were then divided into classes for training. Classes with better scores can be selected as the best way to predict fraud. Therefore, the following feedback strategy is adopted to solve the drift law problem. In this article, we use the European credit card fraud dataset.
Keywords: Credit Card Fraud Detection; Machine Learning Algorithms; Vague Search; See Instructions; Unsupervised Learning; Kev Faib Algorithm.
DOI: https://doi.org/10.38124/ijisrt/IJISRT24MAR388
PDF: https://ijirst.demo4.arinfotech.co/assets/upload/files/IJISRT24MAR388.pdf
REFERENCES