Classifying IoT security risks using Deep Learning algorithms
Wissam Abbass, Zineb Bakraouy, Amine Baïna, Mostafa Bellafkih · 2018
The heterogeneous structure of the Internet of Things (IoT) confronts it to a permanent uncertainty. In fact, one single attack can easily jeopardize its global performance. Therefore, in order to address this problem, we advocate bridging Deep Learning algorithms into the IoT Security Risk Assessment (SRA). Our contribution conveys a Convolutional Neural Network (CNN) model that enhances the performance IoT Security Risk Assessment. Moreover, examination of the provided model usefulness within an experimental case study is entailed. The main contributions of the paper consist on: a novel Deep Learning model for intelligent SRA; Classification of the security risk factors within the IoT and an Evaluation of the proposed model in term of performance and accuracy.As a result, the findings confirm that Deep Learning applied to Security Risk Assessment shows a strong performance optimization.