DDoS Attack Detection Using C5.0 Machine Learning Algorithm

Abstract

Distributed Denial of Service has always been an issue while dealing with network security. The potential of DDoS attacks is not limited by any security measures. This type of attack does not attempt to breach a security perimeter but aims to make the service unavailable to legitimate users. This is particularly an issue in private clouds as public clouds have sophisticated systems to prevent DDoS attacks. DDoS attacks can be used as a shield for other malicious activities. Open resource access model of the Internet is exploited by Distributed Denial of Service attackers. The main objective of this paper is to detect DDoS attacks using C5.0 machine learning algorithm and compare the results with other state of the art classifiers like Naïve Bayes classifier and C4.5 decision tree classifier. The focus is on an offline detection model.