Application-Layer DDoS Detection via Efficient Machine Learning and Feature Selection
Dyari Mohammed Sharif · 2023
Distributed Denial of Service (DDoS) attacks continue to pose a significant threat in the digital landscape, requiring innovative detection strategies to counter their evolving sophistication. In this study, the application of Random Forest, an ensemble machine learning algorithm, coupled with Recursive Feature Elimination as an efficient feature selection technique, is explored for DDoS detection. The aim is to enhance the accuracy and efficiency of DDoS detection systems, incorporating a comprehensive overview of the proposed approach and emphasizing the importance of selective feature inclusion and the synergy between random forest and recursive feature elimination. The proposed approach encompasses dataset acquisition, data preparation, feature selection, classification, and model evaluation, offering a systematic methodology for bolstering DDoS detection capabilities with 99.9% accuracy, 99.9% precision, 99.9% recall, and 99.9% F1 score. The experiments demonstrate that the feature selection strategy significantly reduces computational overhead while maintaining high classification performance.