<![CDATA[The rapid evolution of cyber threats in recent years has intensified the need for intelligent and adaptive security measures. Machine learning (ML) has emerged as a promising solution, offering capabilities for real-time threat detection, prediction, and autonomous response. This systematic literature review aims to investigate the effectiveness of various machine learning algorithms in enhancing cybersecurity between 2018 and 2025. Using a predefined search strategy, articles were sourced from reputable databases including MDPI, ScienceDirect, IEEE Xplore, and SpringerLink. The review focused on peer-reviewed research examining the application of ML in cybersecurity contexts such as threat detection, cyber resilience, and automated incident response. A total of 25 studies were selected after applying strict inclusion and exclusion criteria. The analysis revealed that deep learning and ensemble methods showed superior performance in detecting complex threats, while supervised learning was prevalent in intrusion detection systems. However, issues like data imbalance, adversarial attacks, and ethical transparency were noted as significant challenges. The findings underscore the transformative role of ML in cybersecurity, yet emphasize the need for interpretability and ethical oversight. This review concludes that integrating ML with existing defense systems and human expertise is essential for building adaptive, resilient, and ethical cybersecurity solutions in the evolving digital landscape.]]>
Copyrights © 2025
