A comparative analysis of various machine learning methods for anomaly detection in cyber attacks on IoT networks

Abstract

This study explores the growing challenges of cybersecurity in the context of rapidly adopted Internet of Things (IoT) technologies, which have become increasingly susceptible to cyber threats. The widespread utilisation of IoT systems intensifies the complex interactions between devices and amplifies the traffic of data, creating various opportunities for cyber adversaries. Consequently, detecting and mitigating cyberattacks targeting IoT systems has emerged as a critical imperative in the field of cybersecurity. The primary objective of this study is to employ various machine learning methods to detect cyber anomalies within IoT systems and subsequently compare the efficacy of these methods. Comparative analysis encompasses various machine learning techniques, including Support Vector Machine (SVM), Artificial Neural Network (ANN), Decision Tree (DT), Logistic Regression (LR), and k-Nearest Neighbours (k-NN). This technical evaluation aims to provide a nuanced perspective on the contributions of these methods to the classification of cyber attacks in IoT systems. Performance analysis of these methods serves as valuable information for cybersecurity experts, offering guidance in the development of robust protection strategies for the IoT ecosystem. As IoT security continues to gain prominence, the findings of this study are poised to contribute significantly to the refinement of cybersecurity practices and the fortification of IoT environments against potential threats. In this study, we use different machine learning methods to detect anomalies in cyber attacks on IoT systems and compare the performance of these methods. The result shows that the neural network performed better than the other models.