Malware Detection in IoT Devices Using Machine Learning: A Review

A thorough summary of recent studies on malware detection in Internet of Things environments can be found in the literature review section. The necessity for complex and scalable detection techniques is emphasized by examining various approaches, from signature-based to anomaly detection approaches. Furthermore, the discussion comprises the application of supervised, unsupervised, and semi-supervised machine learning algorithms to mitigate IoT malware threats. While unsupervised techniques like clustering algorithms show promise in detecting new threats, supervised learning techniques like support vector machines (SVM) and random forests (RF) effectively identify known malware patterns. Furthermore, semi-supervised learning techniques present a viable way to improve detection abilities by utilizing compactly labeled data.