Mohammed Gharkan

Deep Learning-Based Intrusion Detection in IoT: A Comprehensive Review of Architectures, Attacks, Challenges, and Future Directions

In a comparative review of forty peer-reviewed studies, it is demonstrated that hybrid DL models provide excellent detection performance (99-100% classification accuracy on benchmark datasets) as well as practical viability for deployment with privacy-preserving Federated Learning for large-scale data.