Gradient Boosting Techniques for Analyzing DDoS Attacks in IoT Networks

The detection and Prevention of distributed-denial-of-service (DDoS) Cyberattacks is one of the primary concerns in network security. This kind of attack overloaded target systems with traffic, making then unreachable to authorized
users. In a typical DDoS scenario, compromised devices are leveraged to generate malicious traffic that floods victim
servers, severely disrupting services. This work presents a machine learningbased Network Intrusion Detection System
(NIDS) for effective DDoS detection on CIC-DDoS2019 dataset. The proposed system integrates robust data
preprocessing, adaptive resampling to address class imbalance, and gradient-boosting classifiers to achieve high
detection accuracy. Comprehensive performance metrics and model explainability techniques are used to validate the
system's effectiveness. Experimental results demonstrate that the approach is scalable, efficient, and well-suited for real-world DDoS detection.