Comparative Analysis of Machine Learning Models in Computer Network Intrusion Detection

Network security is a major concern of the modern era. With the rapid development and massive usage of computer networks over the past decade, the vulnerabilities of network systems continue to be of significant concern. Intrusion detection systems are used to monitor networks and identify unauthorized access or malicious traffic over secured networks. The application of machine learning algorithms to the intrusion detection domain could enhance such systems. This paper presents a comparative analysis of selected machine learning algorithms for network intrusion detection. The CICIDS 2017 dataset provided the necessary dataset for training the models. Results show that of all six considered, Decision Tree classifier was the overall best.