Developing a Network Attack Detection System Using Deep Learning

With the development of networks, the number of network attacks is increasing exponentially. The need for a network security system is becoming more and more important since there is a lot of sensitive information being stored and sent through the Internet. We need to find the best possible ways to protect our systems from any abnormal behaviors that attempt to violate the integrity, confidentiality or availability of useful information. Many different techniques have been used either to prevent or to detect attacks. In this paper, a Deep Learning (DL) approach is used that can create a better and more effective Intrusion Detection System (IDS). The intended approach is based on classifying normal behavior on the network from anomaly behavior. The proposed approach outperforms all the classical approaches with an accuracy of 99% for training and 91.28% for the testing phase, demonstrating its potential for real-time and practical applications.