The Enhanced Anomaly Deduction Techniques for Detecting Redundant Data in IoT

Abstract

Anomaly detection in the Internet of Things presents a significant challenge and is being addressed across various domains, including fraudulent detection, malware protection, information security, and disease diagnosis. However, traditional anomaly detection techniques face limitations when applied directly to IoT due to the distributed nature of wireless transmission and the resource constraints of end nodes. Extracting uncommon behaviors or patterns from complex data in IoT environments poses a difficult task. Therefore, this paper conducts a comprehensive analysis of machine learning-based methods for identifying anomalies in IoT healthcare data. Furthermore, it offers a detailed comparison of their performance, highlighting their respective benefits and disadvantages.