Building a Smarter Shield: Using Ensemble Learning for Multi-Class DDoS Attacks
M. Vijayalakshmi, Arvind Srinivas, S Ramanathan · 2024
Nowadays in this increasing population IBD has become a viral topic that is mainly discussing and implementing topics in both digital and intelligent industry fields. The security problems that widely occur in the signal processing of data streams is a most challenging problem in industries, detection of DDoS attacks in intelligent industrial applications is also highly dangerous. To ensure the security networks, DDoS attack detection is commonly used. Some are failed to detect unknown attacks; they are classical detection that are commonly known as signature-based attacks or explicit-behavior-based attacks. These are very hard to fulfill the requirements of SD-IOT networks. To detect DDoS attacks, the two machine learning algorithms implemented in this paper are Extra Tree and Random Forest. Extratree performed well. Random Forest’s Accuracy is 95.15% and Extratree is 95.59%. To get this accuracy we used UNSW_NB15 as a Dataset. After a target variable has been identified, the data will be divided into two sets: one for training and another for testing. Further it is applied to classification methods. Finally, apart from accuracy precision, recall, and F1-measure is found out.