This paper proposes an intelligent and compact machine learning model for IoT intrusion detection using an ensemble of semi-parametric models with Ada boost. The proposed model provides an adequate real- time intrusion detection at an affordable computational complexity suitable for the IoT edge networks. The proposed model is evaluated against other comparable models using the benchmark data on IoT-IDS and shows comparable performance with reduced computations as required.
|Number of pages||13|
|Journal||International Journal of Computer Networks and Communications|
|Publication status||Published - 2018|
- Adaboosted ensemble learning
- IoT edge security
- Machine learning for IoT