Hybrid Intrusion Detection System Based on Deep Learning

Abstract

Every day, there are new types of cyber-attacks faced by systems and networks of official and non-official organizations, e-commerce, and even people around the world. Since Deep Learning ( DL) can derive better representations from the data and construct better models, this work proposes an Intrusion Detection System (IDS) based on DL techniques by using the Recurrent Neural Network (RNN) algorithm. Hence, this paper presents the design and implementation of the binary class IDS based on RNNs. The Crow Swarm Optimization (CSO) algorithm has been used to reduce the dataset features. This is necessary as reducing the features means dealing with less data, which reflects positively on the system's accuracy and the implementation time. Using the KDD 99 dataset for benchmarking, the experimental results have illustrated that RNN is very suitable for solving the intrusion detection problem in binary classification methods. Indeed, the obtained results have shown the CSO algorithm's superiority for features selection and reduction, where it produced three selected features with an accuracy rate of (98.34%).