Fig. 3From: Intrusion detection in internet of things using supervised machine learning based on application and transport layer features using UNSW-NB15 data-setFeature importance of full features. Feature importance using RF for a full features and b remaining features after removal of features having importance above 0.05. Top contributing features are, a sbytes, sttl, Sload, smean, and ct_state_ttl and b Dload, dmeansz, dur and dttlBack to article page