CCE Theses and Dissertations
Date of Award
2022
Document Type
Dissertation
Degree Name
Doctor of Philosophy (PhD)
Department
College of Computing and Engineering
Advisor
Wei Li
Committee Member
Ling Wang
Committee Member
Ajoy Kumar
Keywords
algorithms, high data volume, intrusion detection systems, machine learning
Abstract
Intrusion detection systems are tools that detect and remedy the presence of malicious activities. Intrusion detection systems face many challenges in terms of accurate analysis and evaluation. One such challenge is the involvement of many features during analysis, which leads to high data volume and ultimately excessive computational overhead. This research surrounds the development of a new intrusion detection system by employing an entropy-based measure called v-measure to select significant features and reduce dimensionality. After the development of the intrusion detection system, this feature reduction technique was tested on public datasets by applying machine learning classifiers such as Decision Tree, Random Forest, and AdaBoost algorithms. We have compared the results of the features selected with other feature selection techniques for correct classification of attacks. The findings demonstrated dimension and data volume reduction while maintaining low false positive rate, low false negative rate, and high detection rate.
NSUWorks Citation
Eljilani Hmouda. 2022. A Validity-Based Approach for Feature Selection in Intrusion Detection Systems. Doctoral dissertation. Nova Southeastern University. Retrieved from NSUWorks, College of Computing and Engineering. (1171)
https://nsuworks.nova.edu/gscis_etd/1171.