An Improved Intrusion Dectection System using BoT-IoT Dataset

Security concern in IoTs is gaining massive attention from researchers. IoT has the power to transform the world. It is grounded in homes, education, energy, mobiles and many more. Thus, it is widely accepted by the public, which makes IoT devices vulnerable to newer kinds of attacks, ultimately posing a challenge to the professionals in cyber security. Intrusion detection is a way to detect the attacks in the traffic data collected from sensors via the network. Over many years, machine learning techniques have been widely used for intrusion detection purposes. This paper proposes a compact and efficient feed-forward artificial neural network with the eXtreme Gradient Boosting algorithm and obtaining satisfactory results for intrusion detection in IoT.