HYBRID MACHINE LEARNING STRATEGIES FORIOT BOTNET DETECTION
Keywords:
Intelligent Botnet Detection, Internet of Things (IoT), Hybrid Machine Learning, Supervised Learning, Unsupervised Learning, Feature Extraction, Clustering Algorithms, Classification Models, Abnormal Behavior, Security, Real-time Detection, False Positives, IoT Network SecurityAbstract
A hybrid machine learning framework is used in this study to make an intelligent botnet detection system for the Internet of Things (IoT). There are big security risks for IoT networks as they grow and become easier for advanced botnet attacks to break into. Most of the time, IoT settings are too big and complicated for old-fashioned detection methods to work. To solve this issue, we suggest a mixed architecture that includes guided and unguided ways of learning. Using methods like feature extraction, clustering, and classification, the system is able to spot strange behaviors that could be signs of botnet activity. Our tests show that our hybrid method is a good and scalable way to find botnets in real time in IoT networks because it improves detection accuracy and lowers false positives.
