Evaluation of machine learning algorithms on internet of things (IoT) malware opcodes

Abstract

Technological advancements have increased the popularity of the Internet of Things (IoT) devices with applications in different spheres of life. As with all technological devices, it is highly prone to malicious attacks such as malware. Several methods have been developed to mitigate these attacks. One of the methods is the use of opcodes to classify malware. These opcodes are generated from disassembled malware programs. Different supervised machine learning classifiers namely Random Forest, K-Nearest Neighbor, Decision Tree, Naïve Bayes, and Support Vector Machine have been implemented. Results of experiments performed using opcodes generated from 512 IoT benign and malware files showed that the Random Forest classifier performed better than other classifiers with an accuracy of 100%.