Feature Vector Generation with Multi Level Classification using Particle Swarm Optimization Model for Intrusion detection
Year: 2024
Type: article
Abstract: The Network Intrusion Detection System (NIDS) is used to detect malicious activities on a network. Machine Learning (ML) techniques are heavily leveraged in the NIDS for intrusion detection. When it comes to enhancing NIDSs' functionality, feature selection is crucial. This is because intrusion identification uses numerous features, each of which must be processed individually. Therefore, the feature selection method impacts the time required to probe traffic behaviour and enhance accuracy. An Intrusion Detection System (IDS) with a powerful intrusion detection mechanism is highly desirable for preventing network intrusion. Despite all the effort put into them, intrusion detection systems are not very effective because of the frequent false positives they produce. Using a raw dataset with redundancy is a common source of false positives. Feature selection, which can boost intrusion detection performance, is required to fix this problem. In this research, we have used a Multi-Level Classification model using Feature Ranking Strategy to perform feature selection (FS) with the goal of eliminating superfluous features. The underlying process of intrusion detection is improved as a result. The false alarm rate was reduced, the detection rate was increased, and the accuracy of the IDS was improved when the Particle Swarm Optimization (PSO) algorithm was used to the selectable features of the NSL-KDD dataset. In order to manage numerous types of attacks, a Rational Multi Level Classification with Feature Ranking Strategy using PSO (RMLC-FRS-PSO) model is designed for accurate detection of intrusions in the network. The proposed model when contrasted with the existing models reduces the false alarms and enhances the network performance.
Author: Battini Sujatha
Language: English
Subfield: Computer Networks and Communications
Field: Computer Science
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W7216855684
