Handling imbalanced data in intrusion detection using time weighted Adaboost support vector machine classifier and crossover boosted Dwarf Mongoose Optimization algorithm

Cybersecurity threats pose a serious challenge in the present day and age, and Intrusion Detection Systems (IDS) have emerged as an effective solution to counter these threats. In this paper, a novel IDS is proposed that captures data from the NSL-KDD dataset and are preprocessed. The Kernel Princip...

Full description

Saved in:
Bibliographic Details
Published in:Applied soft computing Vol. 167; p. 112327
Main Authors: Chandrasekaran, Hemalatha, Murugesan, Kanipriya, Mana, Suja Cherukullapurath, Barathi, Bhagavathi Kannu Uma Anu, Ramaswamy, Sumathy
Format: Journal Article
Language:English
Published: Elsevier B.V 01.12.2024
Subjects:
AdaBoost
Attacks, crossover
Dwarf mongoose
Imbalanced Data
Intrusion detection
Support vector machine and global exploitation
Dwarf mongoose
AdaBoost
Support vector machine and global exploitation
Attacks, crossover
Intrusion detection
Imbalanced Data
ISSN:1568-4946
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Cybersecurity threats pose a serious challenge in the present day and age, and Intrusion Detection Systems (IDS) have emerged as an effective solution to counter these threats. In this paper, a novel IDS is proposed that captures data from the NSL-KDD dataset and are preprocessed. The Kernel Principal Component Analysis (KPCA) model extracts features presented in the data, and the Crossover Boosted Dwarf Mongoose Optimization (CDMO) algorithm selects the relevant features for classification. The CDMO algorithm offers the advantages of improving exploitation, providing optimal solutions, and balancing global exploitation and local search capabilities. The selected features are classified into five classes using the Time Weighted Adaboost Support Vector Machine (TWASVM) classifier. The TWASVM classifier effectively handles imbalanced data and delivers high-performance results. Experiments conducted on MATLAB R2019a and the proposed model achieved an higher accuracy of 98.6 % and less time complexity of 13 seconds. Comparative analysis demonstrated that the proposed IDS outperforms other state-of-the-art methods. The advantages of CDMO algorithm include improved exploitation, optimal solutions, and a balanced crossover strategy for global exploitation and local search capabilities. The advantages of the TWASVM classifier include the ability to handle imbalanced data and deliver high-performance results. Overall, the proposed IDS offer a novel solution to the challenges of intrusion detection in a rapidly evolving cybersecurity landscape. •Novel IDS using NSL-KDD dataset with KPCA and CDMO for feature selection.•CDMO enhances feature selection with improved exploitation and optimal solutions.•TWASVM classifier handles imbalanced data with 98.6 % accuracy and low complexity.•MATLAB R2019a experiments show superior performance over state-of-the-art methods.•IDS provides an advanced solution for modern cybersecurity intrusion detection.
ISSN:1568-4946
DOI:10.1016/j.asoc.2024.112327