A Novel Approach for Intrusion Detection Based on Deep Belief Network

Abstract

In recent years, cyber-attacks have many new forms with larger scale and much more complexity. This requires many network protection solutions and amplify the need for robust cybersecurity practices. One of the effective method to prevent network attacks is to use to Intrusion Detection Systems (IDSs), which can detect attacks never seen before. Many researchers have tried to produce anomaly - based IDSs, but they have not been able to detect malicious network traffic enough to ensure implementation in real networks. Clearly, it is still a challenge for the security community produce IDS suitable for implementation in real world. In this paper, we propose a new approach using a Deep Belief Network called Workflow-based Collaborative Learning with a combination of supervised and unsupervised machine learning methods for network attacks detection. Our proposed approach will be tested with network security datasets and compared with previously existing methods. The experimental evaluation shows that the valid of our approach.