Detection of botnet in IoT network through machine learning based optimized feature importance via ensemble models

The number of cyberattacks has grown along with the expansion of the Internet of Things (IoT), which necessitates detection of cyberattacks on IoT devices. Different machine learning (ML) algorithms, such as Random Forest (RF), Decision Tree (DT), Gradient Boost and novel Voting Ensemble (VE) models are used in this research for botnet detection. This research's goal is to first determine accuracy using a variety of machine learning models, then to apply feature importance for increased efficiency, and finally evaluate the outcomes using novel ensemble models. The efficiency of proposed ensemble models was discovered to be highest after feature importance was applied.