It is difficult to trace IoT Botnet attacks due to their rapidity, out break nature, and limited scale. Machine learning is considered to
be an effective tool for detecting and preventing botnet attacks. Our
work presents a framework for detecting IoT botnets.The function of
machine learning techniques is discussed in this review, where the per formance of seven single classifiers were compared for accuracy, pre cision, recall and other parameters using the CTU13 dataset. Three
ensemble-based classifiers were proposed as well and compared to the
RandomForest classifier. The results exhibit that the RandomForest
classifier had higher accuracy in detecting botnets over the other sin gle classifiers and the ensemble ones. RandomForest classifier demon strated an everage accuracy of 99.91%.
