Machine learning and deep learning for user authentication and authorization in cybersecurity: A state-of-the-art review

In the continuously developing field of cyber security, user authentication and authorization play a vital role in protecting personal information and digital assets from unauthorized use. As the field of cyber security expands, traditional user authentication and authorization approaches are not enough to prevent unauthorized access to personal information. Therefore, Machine Learning and Deep Learning models are introduced in cybersecurity. To assist researchers and cybersecurity experts in their research endeavours, a comprehensive and informative study is required covering the state-of-the-art advancements. Therefore, this research aimed to explore the field of Machine Learning and Deep Learning-based user authentication and authorization. More specifically, this paper intends to explore the diverse application domains of Machine Learning and Deep Learning-based user authentication and authorization. The paper also analyzes the commonly used datasets, pre-processing methods and Machine Learning and Deep Learning algorithms in user authentication and authorization. After that, this study conducts a thorough and detailed examination of some state-of-the-art articles' results and experimental details to enhance comprehension of the present advancements. Finally, the study engages in a comprehensive discussion concerning the various challenges encountered and outlines potential avenues for future research. This systematic review provides an all-encompassing overview of Machine Learning and Deep Learning-based user authentication and authorization, covering its application domains, models, analysis of state-of-the-art results, challenges, and research directions. It serves as a valuable resource for interdisciplinary studies.