VEM-IDSs: enhancing an ensemble learning method of voting approach for binary classification of intrusion detection system

In the realm of cybersecurity, the responsibility for protecting information and data for the Internet of Things is significant, for the prevention of intrusiors from using the data by detecting the vulnerabilities, improve the mechanism, and developing strategies for protecting IoT networks to repel attacks. Intrusion Detection Systems (IDSs) are one of the strategies that can monitor the network to protect data. Machine Learning algorithms are used to increase IDS efficiency and achieve high security. Our study: conducts experiments using ensemble voting (Soft and Hard) classification to detect and classify data in IoT. Our experiments contain two stages. First, build models for three algorithms (K-Nearest Neighbor (KNN), Logistic Regression (LR), and Random Forest (RF). Second, building an ensemble voting classifier consisting of three algorithms. For the two stages, we proposed Edge-IIoTset dataset to classify attacks as benign or malicious, (RF_LR) models result of soft voting have a high accuracy.