Evaluating DDoS Attack Detection Efficiency with Machine Learning Techniques

Abstract

Distributed Denial of Service (DDoS) attacks pose a significant threat to the stability and security of network systems, necessitating robust detection mechanisms. This study evaluates the efficiency of various machine learning techniques in detecting DDoS attacks. We employ a comprehensive dataset containing both legitimate and attack traffic to train and test multiple machines learning algorithms, including Decision Trees, Random Forests, Support Vector Machines, Gaussian Naive Bayes, Logistic Regression and K-Nearest Neighbours. Our evaluation criteria include detection accuracy, false positive rate, and computational efficiency. The results demonstrate that machine learning techniques can significantly enhance the detection of DDoS attacks, with some algorithms outperforming others in specific metrics. Random Forests achieved the highest detection accuracy, while Logistic Regression & Gaussian Naive Bayes offered a balanced trade-off between accuracy and computational cost. The findings underscore the potential of integrating machine learning models into network security infrastructures to pre-emptively identify and mitigate DDoS threats. Future work will explore the adaptability of these models to evolving attack patterns and the integration of real-time detection capabilities.