Deep Learning-Based Approaches for Effective IoT Malware Classification

The Internet of Things (IoT) has gained significant attention as a research domain due to its wide-ranging applications. In this paper , we propose two deep learning models for IoT malware classification. Both models are evaluated through two experiments: the first experiment involves the utilization of chi-square feature selection, while the second experiment incorporates Natural Language Processing (NLP) techniques without feature selection. Our proposed approaches for malware detection achieve a detection accuracy of 87% using Convolutional Neural Networks (CNN) with chi-square feature selection, 89% accuracy using Recurrent Neural Networks (RNN) with chi-square feature selection, 96% accuracy using CNN with NLP techniques, and 99% accuracy using RNN with NLP techniques. Furthermore, the performance of the RNN model with NLP techniques is compared to recent advancements in the field, highlighting its superior effectiveness.