Guest Editor(s)
Prof. Mourade Azrour
Email: mo.azrour@umi.ac.ma
Affiliation: Department of computer sciences, Moulay Ismail University, Meknes, Morocco
Homepage:
Research Interests: Artificial Intelligence (AI), cybersecurity, authentication, intrusion detection
Prof. Abdulatif Alabdulatif
Email: ab.alabdulatif@qu.edu.sa
Affiliation: Department of Computer Science, Qassim University, Buraydah, Saudi Arabia
Homepage:
Research Interests: cybersecurity, authentication
Prof. Azidine Guezzaz
Email: a.guezzaz@uca.ma
Affiliation: Department of Mathematics and Computer Sciences, EST, Cadi Ayyad University, Marrakesh, Morocco
Homepage:
Research Interests: cybersecurity, Artificial Intelligence (AI)
Summary
Anomaly detection systems in the context of network security are designed to identify unusual or deviant patterns in traffic that deviate from normal, established behavior. By focusing on deviations rather than predefined signatures, these systems aim to detect novel or evolving threat activity that may not yet be cataloged in known attack databases. Their importance lies in providing early warning signs and scalable monitoring for modern networks, where traditional rule-based approaches can struggle to keep up with rapidly changing attack vectors and high data volumes.
Key challenges include class imbalance, where malicious events are far rarer than normal traffic, and the representation of diverse or evolving attack types that can be sparsely sampled in training data. Modern anomaly detection leverages ML/AI to automatically extract informative features, model complex patterns across multiple layers, and deliver anomaly scores that help prioritize investigation. This approach enables more robust intrusion detection, supports online learning and adaptation, and can be enhanced with unsupervised or semi-supervised techniques to cope with limited labeled data while maintaining performance in real-world, large-scale environments.
The scope of the Special issue are , but not limited to :
- AI and machine learning for anamaly detection systems
- Deep learning approaches for anomaly detection and classification
- Adversarial AI and its integration for cybersecurity
- AI-driven behavioral analysis for anomaly detection
- Privacy-preserving AI techniques in threat intelligence
- Explainable AI (XAI) for enhanced cybersecurity
- AI applications in cloud, IoT, and edge security
Keywords
intrusion, machine learning, deep learning, anomaly
Published Papers
Login
Register
Submit a Paper
Propose a Special lssue