Ensemble Learning for Detecting Application-Layer DDoS Attacks from Open-Source Toolkits

Abstract

The aim of the project is to detect and alleviate DDoS Application-Layer escalating attacks and provide insight into patterns of attack and tools for increased cyber security measures. In order to attack HTTP-birds, the project seeks to detect tactics and tools and offer a specialized approach to strengthen understanding and countermeasures against developing cyber threats. It is urgently necessary to solve the growing threat of DDOS by shifting the project focus on the accessibility of tools. This is essential for active defense against the widespread use of harmful offensive equipment. The aim of the project is to seize the network administrator and cyber security experts and provide online services. Finally, it benefits users and businesses with durable defense against developing DDOS threats. “To boost performance, we introduced ensemble models—voting Classifier (RandomForest, DecisionTree) and Stacking Classifier (RandomForest, DecisionTree, LGBM)”. The aim of these improvements is to improve the accuracy of cyberbully detection.