Applying Machine Learning for Recognition of DDoS Attacks using NSL- KDD

Abstract

DDoS (Distributed Denial of Service) attacks are significantly dangerous. Spotting them is very essential for system availability and performance. DDoS detection is noticing a growing attempt in network traffic and reducing its effects. To identify or detect DDoS attacks, many methods have been developed including Statistical analysis, Machine learning and anomaly detection. To distinguish between legitimate and malicious data the features rely on packet rates, flow patterns and network behavior. The fineness of the DDoS detection depends on the classification methods and the accuracy of the classification algorithm. Attackers continue to create new strategies to avoid detection in DDoS. Therefore, it is very necessary to increase the accuracy and effectiveness of DDoS detection algorithms.