IoT Intrusion Detection Using Machine Learning with a Novel High Performing Feature Selection Method

The Internet of Things (IoT) ecosystem has experienced significant growth in data traffic and consequently high dimensionality. Intrusion Detection Systems (IDSs) are essential self-protective tools against various cyber-attacks. However, IoT IDS systems face challenges due to functional physical diversity. These characteristics make exploiting all features attributes for self-protection difficult unrealistic. This paper proposes implements a novel feature selection extraction approach (i.e., our method) anomaly-based IDS. begins with using two entropy-based approaches information gain (IG) ratio (GR)) select extract relevant ratios. Then, mathematical set theory (union intersection) is used the best features. model framework trained tested on intrusion dataset 2020 (IoTID20) NSL-KDD four machine learning algorithms: Bagging, Multilayer Perception, J48, IBk. Our resulted 11 28 (out 86) intersection union, respectively, IoTID20 15 25 41) NSL-KDD. We have further compared other state-of-the-art studies. comparison reveals that superior competent, scoring very 99.98% classification accuracy.