Hybrid Machine Learning based Intrusion Detection System for IoT
Kajol Mittal, Payal Khurana Batra · 2022 International Conference on Computing, Communication, and Intelligent Systems (ICCCIS) · 2022
The Internet of Things (IoT) is susceptible to several attacks by insiders and outsiders. IoT requires an effective Intrusion Detection System (IDS). The development of IDSs that can quickly and automatically identify and classify cyberattacks at the network and host level commonly uses machine learning techniques. This research proposes a hybrid machine learning model that combines Naive Bayes and K-NN. The UNSW-NB15 dataset, open to the public, is used to test the suggested model. Accuracy, precision, recall, and f1-Score are four well-known performance evaluation metrics used to analyze the model’s performance. The experimental results show that the hybrid approach performs better than the individual ML model, with the accuracy of 91.8% in the case of multiclass classification and 96.2% in the case of binary classification.