Applying Supervised Machine Learning Techniques to Detect DDoS Attacks

Distributed Denial of Service (DDOS) attack is one of the most devastating attacks, since it disrupts the performance of fundamental services provided by many companies on the internet. Many studies have been conducted by researchers to address this issue; nevertheless, owing to the inefficiency of existing methodologies in terms of accuracy and computational cost, presenting an effective method to identify DDoS attacks remains an exciting research topic. As a result, DDoS attack detection research is now becoming significantly important. In this paper, we used various machine-learning methods to detect DDoS attacks. Several classifiers such as Random Forest (RF), K- nearest neighbor (KNN), Support Vector Machine (SVM), and Decision Tree (DT) were implemented, and their performance was measured on the UNSW-NB15 benchmark dataset. The experimental results reveal that the Random Forest model performed well compared to state- of-the-art approaches, with an accuracy of 94.14 percent on train data and 91.88 percent on the entire test dataset. Also, The RF model can more reliably discriminate between malicious and benign traffic, with a greater detection rate and a lower false alarm rate.