Метод виявлення аномальної мережевої поведінки IoT-пристроїв з використанням методів машинного навчання

Abstract

The growth of IoT devices has increased network attack surfaces and exposed the limitations of signature-based intrusion detection systems. This study explores machine learning methods for anomaly detection in IoT traffic using behavioural network features extracted from benign and malicious datasets. The findings show that models such as Isolation Forest effectively detect abnormal activity and are suitable for real-time monitoring applications.