A Novel SVM Based CFS for Intrusion Detection in IoT Network

Abstract

Due to the rapid growth of technologies, the internet of things become an important research topic, which can ensure collecting and transferring data over the network between connected objects without any human intervention. However, these connected objects are generally constrained by energy consumption and data security in terms of confidentiality, integrity and availability against attackers. This paper presents a solution for this problem, in this content, we propose an intelligent host-based intrusion detection system using machine learning. our approache based on Support vector machine (SVM) is implemented, we used correlation-based feature selection (CFS) technique to detect the pertinent features in the NSL-KDD dataset. The experimental results show that our approach has an accuracy as 99.09% in binary classification and 99.11% in multiclass classification which are performed better than most of previous approaches.