| Abstract: |
The rapid proliferation of cyber threats and sophisticated attack vectors in modern networked environments has necessitated the development of intelligent, adaptive defense mechanisms beyond the capabilities of traditional rule-based security systems. Machine learning (ML) and deep learning (DL) have emerged as transformative paradigms in cybersecurity, offering unprecedented capacities for threat detection, anomaly identification, malware classification, and intrusion prevention. This review paper presents a comprehensive meta-analysis of the existing literature on the application of ML and DL techniques in cybersecurity domains published over the past decade. Through systematic analysis of over 150 peer-reviewed studies, the paper synthesizes key findings across sub-domains including network intrusion detection systems (NIDS), malware analysis, phishing detection, vulnerability assessment, and adversarial robustness. The review critically evaluates the performance of widely adopted algorithms such as Random Forest, Support Vector Machines, Convolutional Neural Networks (CNN), Recurrent Neural Networks (RNN), Long Short-Term Memory (LSTM), and Generative Adversarial Networks (GAN) across benchmark datasets including NSL-KDD, CICIDS2017, UNSW-NB15, and DREBIN. Methodological gaps, class imbalance issues, adversarial vulnerabilities, and real-world deployment challenges are identified. |
