Improving IoT Botnet Detection Using Ensemble Learning

With the increasing use of Internet of Things (IoT) devices in various domains, including offices, homes, hospitals, cities, and transportation, cyberattacks using malicious attacks have become more frequent and complex, posing new challenges and risks. Therefore, it is crucial to enhance the speed and accuracy of security measures. In this paper, we propose an ensemble machine-learning model that utilizes various techniques, such as Stacking and Bagging, in combination with individual classifiers based on machine learning models to detect botnet attacks using the N-BaIoT dataset. Our results demonstrate the efficiency and efficacy of the proposed stacking model, which outperformed other techniques for every evaluation metric. We conclude that the selected model can achieve a very good accuracy rate.