A review of machine and deep learning techniques for network intrusion detection

The rapid development of the Internet and communication technology has led to the expansion of large networks and data. In response to these threats, intrusion detection systems (IDS) were created to protect networks by analysing network traffic to ensure privacy, fairness, and security. The challenge remains correcting, reducing, and identifying new inputs. Recently, IDS based on machine learning (ML) and deep learning (DL) have been offered as effective techniques for detecting network vulnerabilities. This paper provides an overview of IDS and then classifies important ML and DL techniques for developing network-based IDS (NIDS) systems. Furthermore, this paper updates the techniques, evaluation methods, and data selection to reflect the current needs and advances in ML and DL-based NIDS. The limitations of the proposed method are analysed, key research issues are identified, and future research plans for NIDS based on ML as well as DL are provided.