A State-of-the-Art Review of Machine Learning in Cybersecurity Data Science

With the proliferation of cybercrime in computerized systems, the demand for effective cybersecurity measures has escalated. Data science, powered by machine learning (ML) and deep learning (DL) techniques, has emerged as a critical tool for detecting and mitigating cyber threats. In this paper, we conduct a comprehensive survey of recent literature in cybersecurity data science, focusing on ML and DL methodologies. We review over 50 papers and analyze various types of cyberattacks, including denial-of-service, zero-day attacks, phishing, and insider threats. Additionally, we propose a framework for future research in cybersecurity. Our comparison analysis encompasses malware detection, intrusion detection, anomaly detection, and Internet of Things (IoT)-based technologies. Despite the challenges posed by evolving cyber threats, our survey demonstrates the efficacy of supervised and unsupervised DL strategies in categorizing cyberattacks and informs decision-making processes in cybersecurity defense.