Hybridized Model of an Improved Genetic Algorithm with Local Linear Embedding Algorithm for Feature Selection in Intrusion Detection Systems

Simple item page
dc.contributor.authorTolulope Olushola OLUFEMI
dc.date.accessioned2024-07-01T11:57:32Z
dc.date.available2024-07-01T11:57:32Z
dc.date.issued2023-12
dc.description.abstractAbstract The Internet has revolutionized various sectors, offering opportunities for innovation and advancement. However, it brings the risk of cyber-attacks. Intrusion Detection Systems (IDS) are vital in identifying and preventing such attacks. The quality of IDS models relies on selecting relevant features during the training process. This study proposes a hybrid optimization model for feature selection in IDS to enhance the efficiency and accuracy of the system, and the methodology incorporates an improved Genetic Algorithm (I-GA) and Local Linear Embedding (LLE) optimization techniques. It aims to identify the most relevant features for building an appropriate and effective IDS model. The Machine learning classifiers, including Support Vector Machine (SVM) and k-nearest Neighbors (KNN), use the reduced and highly relevant features obtained through the I-GA with LLE approach for training and testing purposes. The performance of these combinations is validated using various metrics, and the results demonstrate the effectiveness of the proposed approach. The accuracy achieved for I-GA-LLE-SVM, I-GA-LLE-KNN, the LLE-SVM, the LLE-KNN, the I-GA-SVM, I-GA-KNN are 94%, 99%,86%,98%,89% and 98% respectively. These results highlight the I-GA-LLE approach's superiority over the other combinations. The significant improvements in accuracy obtained through the I-GA-LLE-SVM and I-GA-LLE-KNN combinations emphasize the importance of integrating multiple optimization techniques for feature selection in IDS models. In conclusion, this study presents a hybrid optimization model that effectively selects relevant features for IDS. The combination demonstrates superior performance in terms of accuracy. The findings emphasize incorporating advanced optimization techniques and multiple classifiers for outstanding intrusion detection, which can significantly enhance cybersecurity measures and safeguard networks in the face of evolving cyber threats. Keywords: Intrusion Detection System, Support vector machine, k-nearest Neighbor, Machine Learning, Genetic Algorithm Word Count: 275
dc.identifier.citationKate Turabian
dc.identifier.otherPh.D
dc.identifier.urihttps://repository.lcu.edu.ng/handle/123456789/633
dc.language.isoen
dc.publisherLead City University
dc.relation.ispartofseriesPh.D
dc.subjectIntrusion Detection System
dc.subjectSupport vector machine
dc.subjectk-nearest Neighbor
dc.subjectMachine Learning
dc.subjectGenetic Algorithm
dc.titleHybridized Model of an Improved Genetic Algorithm with Local Linear Embedding Algorithm for Feature Selection in Intrusion Detection Systems
dc.typeThesis

Files

Original bundle
Now showing 1 - 2 of 2
No Thumbnail Available
Name:
Olufemi Tolulope Olushola Full Thesis.pdf
Size:
28.36 MB
Format:
Adobe Portable Document Format
Download
No Thumbnail Available
Name:
Olufemi Tolulope Olushola Preliminary Pages.pdf
Size:
2.52 MB
Format:
Adobe Portable Document Format
Download
License bundle
Now showing 1 - 1 of 1
No Thumbnail Available
Name:
license.txt
Size:
1.71 KB
Format:
Item-specific license agreed to upon submission
Description:
Download

Collections

Department of Computer Science;