Article: Robust intrusion detection techniques on improved features selection generalised variable precision rough set Journal: International Journal of Information and Computer Security (IJICS) 2024 Vol.25 No.1/2 pp.116 - 140 Abstract: Network-based communication is becoming more and more susceptible as it is used extensively for outsiders and attacks in many areas. Intrusion detection is an essential process for a complete communication network security strategy. Intruders learn tactics of attacks every day, so they try to observe the significance of the intrusion detection system thoroughly, and they deny the services of IDS to the respective users. The three prominent roles that perform essential tasks in the network security of IDS are data collection, selection of optimal parameters, and classification made by decision-making engines. The recent research area highly relies on selecting an IDS optimal feature. Machine learning has explored various novel methods to improve performance and achieve a high accuracy rate. The proposed work implements a generalised rough set theory for optimal parameter selection, which leads to a formal way to enhance the accuracy. Support vector machines are used to classify network packet threats using machine learning. The suggested work uses the NSL-KDD dataset because it improves network communication security. Pre-processing data and feature selection on generic variable precision rough sets should be compared to best initial search and genetic search. Inderscience Publishers - linking academia, business and industry through research

Title: Robust intrusion detection techniques on improved features selection generalised variable precision rough set

Authors: R. Rajeshwari; M.P. Anuradha

Addresses: Department of Computer Applications, Bishop Heber College, India; Affiliated to: Bharathidasan University, India ' Department of Computer Science, Bishop Heber College, India; Affiliated to: Bharathidasan University, India

Abstract: Network-based communication is becoming more and more susceptible as it is used extensively for outsiders and attacks in many areas. Intrusion detection is an essential process for a complete communication network security strategy. Intruders learn tactics of attacks every day, so they try to observe the significance of the intrusion detection system thoroughly, and they deny the services of IDS to the respective users. The three prominent roles that perform essential tasks in the network security of IDS are data collection, selection of optimal parameters, and classification made by decision-making engines. The recent research area highly relies on selecting an IDS optimal feature. Machine learning has explored various novel methods to improve performance and achieve a high accuracy rate. The proposed work implements a generalised rough set theory for optimal parameter selection, which leads to a formal way to enhance the accuracy. Support vector machines are used to classify network packet threats using machine learning. The suggested work uses the NSL-KDD dataset because it improves network communication security. Pre-processing data and feature selection on generic variable precision rough sets should be compared to best initial search and genetic search.

Keywords: intrusion detection system; IDS; anomaly detection; generalised variable precision rough set; GVPRS; feature selection; machine learning; support vector machine; SVM; NSL-KDD dataset.

DOI: 10.1504/IJICS.2024.142689

International Journal of Information and Computer Security, 2024 Vol.25 No.1/2, pp.116 - 140

Received: 29 Jul 2023
Accepted: 23 Nov 2023
Published online: 18 Nov 2024 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article

Title: Robust intrusion detection techniques on improved features selection generalised variable precision rough set

Keep up-to-date