Experimental results show better performance of the proposed hybrid IDS using an Arithmetic Optimizer Algorithm (AOA) and Majority Vote Classifier (MVC) in terms of higher intrusion detection accuracy and fewer features compared to other similar studies.
Intrusion Detection System (IDS) has been an imperative challenge in Computer Networks. Commonly, based on the network traffic and large amount of transmitted data in the network, solutions preventing misdiagnosis of attacks and increasing the accuracy of intrusion detection has been proposed by researchers. To address the mentioned challenge, in this paper we propose a hybrid IDS using an Arithmetic Optimizer Algorithm (AOA) and Majority Vote Classifier (MVC) for computer networks. First, the optimal feature subset is selected by the Arithmetic Optimizer Algorithm and then a MVC is used to classify the samples. MVC utilizes Naive Bayes (NB), Decision Tree (DT), and k-nearest neighbors (KNN). The efficiency of the proposed method has been evaluated using the UNSW-NB15 dataset and the results have been compared with other methods such as Particle Swarm Optimization (PSO), Genetic Algorithm (GA), and Differential Evolution (DE) as well as similar methods. Experimental results show better performance of the proposed method in terms of higher intrusion detection accuracy and fewer features compared to other similar studies.