Detection Of DDOS Attack using Machine Learning Models
S. Santhosh, M. Sambath, J. Thangakumar
Abstract
The most dangerous cyberattack is the distributed denial-of-service (DDoS) attack, which is a more sophisticated version of the denial-of-service (DoS) attack. Like DoS, DDoS tries to prevent users from accessing system services. DDoS attacks differ from DoS attacks in that they originate from multiple hosts as opposed to a single host. which adversely impacts the targeted system. Therefore, a good model needs to be framed to identify DDoS incidents. Even though many researches has been done. Due to the nature of DDoS attack mechanisms, the appropriate approach is not framed till now with high detection rate. In this project, a machine learning model-based approach to DDoS attack detection is proposed. The attack detection model is trained and tested using the XGBoost Classifer and Random Forest, the results of the approaches are to be visualised and compared in order to determine the model with the best accuracy rate. The XGBoost algorithm is modified to achieve a higher accuracy rate.