An Enhanced Approach to DDoS Detection using Machine Learning models with Explainability
R. Vaishali, Vijayakanthan Ganesalingam · 2024
The widespread use of the Internet for various purposes has increased the vulnerability to cyberattacks, particularly Distributed Denial of Service attacks, which severely disrupt system functionality by interrupting communication. These attacks have led to significant losses, prompting numerous studies aimed at mitigating their impact and protecting network systems. This study presents a comparative analysis of several machine learning models using a dataset specifically developed for DDoS detection. Among the evaluated algorithms, the CatBoost classifier with optimized parameters demonstrated superior performance, achieving an accuracy of 99.99%. This research further examines the CatBoost model in comparison with other algorithms and discusses its interpretability using explainable AI (XAI) technique.