Refine Framework for Credit Card Fraud Detection System using Machine Learning

Abstract

In today's digital landscape, many systems exist to detect credit card fraud. This study seeks to investigate the various methodologies employed in Credit Card Fraud Detection, as well as the selection and pre-processing of datasets necessary for developing Machine Learning, Deep Learning, and Neural Network models. The research will encompass a range of models, including decision trees, logistic regression, neural networks, Gaussian kernels, mining-based neural networks, self-organizing maps, generative adversarial networks, ensemble learning techniques, AdaBoost, majority voting, deep convolutional neural networks, adversarial learning, fuzzy clustering, optimized light gradient boosting, anti-k nearest neighbor methods, calibrated probabilities, and bidirectional Long Short-Term Memory (BiLSTM) and Gated Recurrent Neural Networks.