Застосування методів машинного навчання для виявлення аномалій та кібератак в мережах інтернету речей (IoT)

Abstract

The exponential growth of the Internet of Things (IoT) has introduced unprecedented connectivity, but it has also significantly expanded the attack surface for cyber threats. This paper explores the critical role of Machine Learning (ML) in enhancing IoT cybersecurity through advanced anomaly detection. Unlike traditional rule-based systems, ML algorithms dynamically adapt to identify zero-day attacks and deviations from baseline behavior. The study reviews both supervised and unsupervised learning techniques, highlighting the efficacy of models such as Isolation Forest and Autoencoders in resource constrained environments. Ultimately, the paper concludes that integrating intelligent, data-driven approaches is essential for building resilient, autonomous security frameworks in modern cyber-physical systems.