A Comprehensive Analysis of the Machine Learning Algorithms in IoT IDS Systems

Abstract

In this study, machine learning algorithms in IoT IDS (Internet of Things Intrusion Detection System) systems are comprehensively compared from various aspects. Accuracy, precision, and training time are evaluated. The effects of data preprocessing techniques including normalization, outlier removal, standardization, and regularization on the datasets are examined. Furthermore, the impact of dataset balancing, considering both balanced and imbalanced scenarios, on machine learning performance is investigated. The contribution of feature selection on the four different datasets is also analyzed. Based on findings, it is observed that certain preprocessing operations provide significant advantages in various ML algorithms, whereas others have very low impact, and their performance varies depending on the dataset and feature selection. The aim of this study is to facilitate the complexity and lengthiness of machine learning processes and algorithm selection, providing insights for future academic research. By addressing this objective, an effort is made to shed light on simplifying the utilization of machine learning algorithms. The challenges arising from the complexity of machine learning processes in IoT IDS systems are addressed by this study. This contribution can greatly benefit researchers in their academic endeavors. This multifaceted approach proves beneficial when comparing the methods under consideration, fostering a scientific discourse on their efficacy within contexts.