Performance Comparison of Intrusion Detection System Using Three Different Machine Learning Algorithms
Zena Khalid Ibrahim, Mohammed Younis Thanon · 2021
Intrusion Detection Systems (IDS) are essential for Network Security in order to control the network and also to analyze the incoming network traffic. With the existence of a huge volume of data and network traffic, there is a need to develop modern technologies like big data and the Internet of Things (IoT), and Cloud Computing (CC). In this article, a comparative study is performed for three Machine Learning algorithms that were implemented on the NSL- KDD dataset for the IDS system. To obtain the optimal accuracy, it is required to select the appropriate set of features in a large dataset. So, the ANOVA F-test and Recursive Feature Elimination (RFE) was used to select important features. The authors have conducted an experiment on IDS that uses three different Machine Learning algorithms, Random Forest (RF), K Nearest Neighbor (KNN), and Support Vector Machine (SVM). The performance of the different models was compared using all the features and the best-selected features were executed using the confusion matrices.