A comprehensive review of AI based intrusion detection system

Title

A comprehensive review of AI based intrusion detection system

Creator

Sowmya T.; Mary Anita E.A.

Description

In today's digital world, the tremendous amount of data poses a significant challenge to cyber security. The complexity of cyber-attacks makes it difficult to develop efficient tools to detect them. Signature-based intrusion detection has been the common method used for detecting attacks and providing security. However, with the emergence of Artificial Intelligence (AI), particularly Machine Learning, Deep Learning and ensemble learning, promising results have been shown in detecting attacks more efficiently. This review discusses how AI-based mechanisms are being used to detect attacks effectively based on relevant research. To provide a broader view, the study presents taxonomy of the existing literature on Machine Learning (ML), Deep learning (DL), and ensemble learning. The analysis includes 72 research papers and considers factors such as the algorithm and performance metrics used for detection. The study reveals that AI-based intrusion detection methods improve accuracy, but researchers have primarily focused on improving performance for detecting attacks rather than individual attack classification. The main objective of the study is to provide an overview of different AI-based mechanisms in intrusion detection and offer deeper insights for future researchers to better understand the challenges of multi-classification of attacks. 2023

Source

Measurement: Sensors, Vol-28

Date

2023-01-01

Publisher

Elsevier Ltd

Subject

Artificial intelligence; Deep learning; Ensemble learning; Intrusion detection system; Machine learning

Coverage

Sowmya T., Department of CSE, Christ University Bangalore, CMR Institute of Technology, Bengaluru, India; Mary Anita E.A., Department of CSE, Christ University, Bangalore, India

Rights

All Open Access; Gold Open Access

Relation

ISSN: 26659174

Format

Online

Language

English

Type

Article

Identifier

https://doi.org/10.1016/j.measen.2023.100827

https://www.scopus.com/inward/record.uri?eid=2-s2.0-85163471347&doi=10.1016%2fj.measen.2023.100827&partnerID=40&md5=38c5ff8dc0c55fc5ce68b9a51dca1a02