Real-Time DDoS Attack Prediction using Supervised Algorithms and CNN

DDoS is an attack that takes an advantage of network user request capacity restrictions, such as the structure that enables a function from a website of a company. The goal of this research is to offer an overview of several DDOS attack detection strategies at the network level. After the completion of a theoretical strategy, quantitative data are incorporated from multiple tested approaches as well as numerous tests are performed in a laboratory environment. The main purpose is to provide a basic understanding of machine learning techniques and how they might be utilized to develop a real-time DDoS attack prediction approach which may be helpful for future academics. Here, 17 features among 33 are used to improve the performance results. In this study, CNN outperforms the other five supervised learning algorithms. Finally, 85.4% accuracy has been achieved when using CNN approaches.