An Intelligent Framework for Smart Automated House Implementation via Integration of IOT and DL

The danger environment for the Internet of Things (IoT) has grown as a result of the widespread use of internet-connected devices, and malware detection is becoming increasingly difficult. Feature extraction, selection, and classification phases of machine learning turn it into a potent weapon against many dangers in Internet of Things environments. Particularly, in order to decrease feature size and speed up detection, feature selection is critical. Three approaches are usually used by researchers: feature reduction, feature selection, or hybrid approaches. This study examines feature selection and feature reduction for IoT malware identification using deep learning (DL) and machine learning (ML) techniques. According to the findings, feature selection by itself can lead to an accuracy rate of $94 \%$. This demonstrates that when it comes to recognizing IoT malware, feature selection works better than feature reduction.