Modelling of Hybrid Intrusion Detection System in Internet of Things with Machine Learning Approach

Abstract

Internet of Things (IoT) devices are interconnected devices that integrate things and the Internet to make human life easy and faster and also Internet of Things devices are interconnected for a longer period without human intervention. This raises to develop security solutions to handle the security issues in the IoT network which is compatible with the services. In this research, we used raw data to construct the model for the system and after the data is prepared there are different mechanisms that we follow to analyze the data; data pre-processing for removing the irrelevant feature in the data, feature selection for selecting features using random forest algorithm. We conduct our experiments by selecting four different supervised machine learning for the classification of attacks on the IoT network. From the experimental result cascading two machine learning algorithms (Random forest and Support Vector Machine) performance is better than among other cascading machine learning algorithms.