In the field of network security, intrusion detection system plays a vital role in the procedure of applying machine learning (ML) techniques with the dataset. This study is an IDS related in machine, developed the literature by utilizing AWID dataset. There tends to be a need in balancing a dataset and its existing approaches from the analysis of its respective works. A taxonomy of balancing technique was introduced due to the lack of treatment of imbalance. This attempt has provided a proper structure defined on all levels and a hierarchical group was formed with the collected papers. This describes a comparative study on the proposed or treated aspects. The main aspect from the surveyed papers were found that: understanding of the existing taxonomies were not in detail and there were no treatment of imbalance for the utilized dataset. So, this study concludes a gathered information in these aspects. Regardless, there are factors or weakness have been seen in any adaptations of the intrusion detection system. In this context, there are few findings that are multifold with contributions. Thus, to best of our knowledge, the study provides an integration with the observation of threshold limit and feature drop selection method by random samples. Thus, the work contributes a better understanding towards imbalanced techniques from the literature surveyed. Hence, this research would benefit for the development of IDS using ML.
Library of Congress Subject Headings
Wireless communication systems--Security measures; Wireless LANs--Security measures; Machine learning; Data sets--Quality control
Networking and System Administration (MS)
Robert Wilson, Dino Linekar, "Towards Effective Wireless Intrusion Detection using AWID Dataset" (2021). Thesis. Rochester Institute of Technology. Accessed from