Paper Contents
Abstract
The increase in instances of credit card fraud has required the creation of advanced detection and categorization systems. This research paper examines the latest breakthroughs in artificial intelligence (AI) models specifically developed to counteract credit card fraud. The study showcases the efficacy of different machine learning (ML) and deep learning (DL) techniques, such as supervised and unsupervised learning methods, anomaly detection, and neural network topologies, by analysing research conducted between 2021 and 2023. The text also covers the incorporation of sophisticated data preprocessing techniques and feature selection approaches.The paper discusses the difficulties encountered in this field, including the unbalanced characteristics of fraud detection datasets and the ever-changing strategies employed by fraudsters. The key findings suggest that the use of hybrid models, which integrate various techniques and utilise ensemble learning, greatly enhances the accuracy of detection and the performance of classification. This text examines the deployment of fraud detection systems that operate in real-time, as well as the significance of interpretability in AI models. It highlights the crucial nature of model transparency and dependability.To summarise, this work proposes potential areas for future research, such as integrating explainable AI (XAI) to improve the transparency of models and employing transfer learning to enhance their adaptability. It is advisable to consider utilising blockchain technology for the purpose of safeguarding transaction data. This review is a significant reference for researchers and practitioners seeking to create or improve AI-powered solutions for detecting credit card scams, ultimately leading to more secure and dependable financial systems.
Copyright
Copyright © 2024 NITESH SHARMA. This is an open access article distributed under the Creative Commons Attribution License.
