Μηχανική μάθηση στην ανίχνευση κυβερνοεπιθέσεων : βιβλιογραφική ανασκόπηση των μεθόδων και των τεχνολογιών
Machine learning in cyberattack detection : a literature review of methods and technologies
View/
Keywords
Κυβερνοασφάλεια ; Μηχανική μάθηση ; Ανίχνευση εισβολών ; Ανίχνευση ανωμαλιών ; Αλγόριθμοι κατηγοριοποίησης ; Βαθιά μάθηση ; Νευρωνικά δίκτυα ; Κυβερνοαπειλές ; Συστήματα IDS ; Μελλοντική έρευναAbstract
The rapid proliferation of cyber threats and the increasing complexity of modern attacks highlight the urgent need for advanced and adaptive detection and mitigation mechanisms. This study explores the contribution of machine learning to cybersecurity enhancement, with a particular focus on cyberattack detection through intelligent algorithms. Initially, the theoretical background of cybersecurity and the core principles of machine learning are presented. The analysis then focuses on the most prominent categories of algorithms applied in intrusion detection, including classification models, neural networks, and anomaly detection techniques. Through an extensive literature review, the performance of these methods is evaluated on representative datasets, and key research approaches in the field are identified. The study concludes with critical findings regarding the effectiveness of these techniques, while also highlighting existing challenges and future research directions. The main objective is to contribute to a deeper understanding of the technological capabilities and limitations of machine learning in the context of cybersecurity, offering a useful reference framework for researchers and professionals in the field.