Machine Learning based Intrusion Detection System for Minority Attacks Classification

Abstract

With the fast usage of internet services, the Internet of Things (IoT) and edge machines are also connecting to the World Wide Web (WWW). The frequent utilization of these devices poses a significant amount of challenge to network security researchers to deal with advanced or sophisticated cyber threats. Traditional based antiviruses and Intrusion detection system (IDS) mechanisms failed to detect advanced cyber threats like minority and majority attacks. Therefore, an efficient mechanism is required to counter this type of cyber-attacks. This paper introduces an efficient IDS mechanism using several supervised machine learning algorithms. The machine learning classifiers are performed on the CIC-IDS dataset 2017 to analyze and inspect the performance of the presented model. The presented method displayed the average accuracy (99%), recall (100%) for four classifiers, namely, Random Forest (RF), Decision Tree (DT), Extra Tree (ET), and K-Nearest Neighbor (K-NN).