Network Intrusion Detection System Using Machine Learning

Abstract

The latest advances in the internet and communication areas have resulted in a massive expansion of network size and data. As a result, plenty of new dangers have arisen, making it difficult for network security to identify attacks effectively Furthermore, intruders with the intent of executing innumerable assaults within the network cannot be overlooked. An intrusion detection system (IDS) is a tool that inspects network traffic to verify confidentiality, integrity, and availability. Despite the researchers' best efforts, IDS continues to encounter difficulties in boosting detection accuracy while lowering false alarm rates and detecting fresh intrusions. Machine learning (ML)-based IDS systems have recently been deployed as promising solutions for quickly detecting intrusions across the network. This article defines IDS and then presents a taxonomy based on prominent machine learning techniques used in the construction of network-based IDS (NIDS) systems. The benefits and drawbacks of the proposed solutions are discussed in depth in this detailed evaluation of current NIDS-based studies. The proposed technique, evaluation criteria, and dataset selection are then discussed, as well as recent trends and breakthroughs in ML-based NIDS. We highlighted many research obstacles and recommended future research scope for improving ML-based NIDS using the weaknesses of the proposed approaches.