Hybrid supervised model with nature inspired optimization for intrusion detection in iot environment

Abstract

IoT data production spans across various domains and applications, such as smart homes, industrial automation, healthcare, agriculture, transportation, and more. The continuous growth of IoT devices and applications generates vast amounts of data, necessitating the extraction of optimal insights to derive valuable knowledge and enable informed decision-making. However, many research studies rely on sampling techniques and feature extraction, often leading to the loss of valuable information from the raw data. Consequently, this results in reduced judgment accuracy and limits the applicability of these approaches in real-world scenarios. Addressing the challenges faced by conventional machine learning algorithms in designing IoT intrusion detection system models, the research aims to investigate the integration of Nature-Inspired Optimization techniques for hyperparameter tuning. This integration will aid in identifying ideal parameters for constructing the IoT IDS model, thereby enhancing its accuracy and effectiveness. newline