Optimising Internet of Medical Things Intrusion Detection Performance using Machine Learning

Doherty, Charles (2025) Optimising Internet of Medical Things Intrusion Detection Performance using Machine Learning. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (1MB) | Preview
Preview	PDF (Configuration Manual) Download (814kB) | Preview

Abstract

Internet of Medical Things (IoMT) networks are prime targets for cyber attacks due to their transmission and storage of sensitive health-related data. Performance advancement of lightweight machine learning (ML) algorithms is crucial to maximise the effectiveness of intrusion detection systems (IDS) in IoMT due to resource constraints. A review of related works noted an absence of detailed studies on how hyperparameter tuning efforts may improve the operation of an IDS specifically in an IoMT environment. Previous studies have been hindered by available datasets lacking diverse attack methods and devices specific to IoMT. This innovative research paired the assessment of multiple hyperparameter tuning methods with a recently published state-of-the-art dataset, the CICIoMT2024 dataset, to investigate potential performance enhancements. Inclusion of comprehensive details of input parameters provided for each tuning method is often omitted from research in the fields of hyperparameter tuning and intrusion detection. Publication of this information in this study enhances reproducibility of the work and benefits future studies. Selected models tested in this study were found to outperform those published by the authors of the CICIoMT2024 dataset for both 2-class (Binary: benign or malicious) and 6-class (Categorical: benign or category of attack) classification. Improvements in balancing performance across all categories was also obtained in results for Categorical classification in comparison to a recent publication. This work may be used as a foundation to improve generalisation of IoMT IDS performance by introducing additional datasets or expanding the number of categories used for multi-class classification efforts.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Spelman, Ross UNSPECIFIED
Uncontrolled Keywords:	Internet of Medical Things (IoMT); Intrusion Detection; Machine Learning (ML); Hyperparameter Tuning; Cybersecurity
Subjects:	Q Science > QA Mathematics > Computer software > Computer Security T Technology > T Technology (General) > Information Technology > Computer software > Computer Security R Medicine > Healthcare Industry T Technology > TK Electrical engineering. Electronics. Nuclear engineering > Telecommunications > Computer networks > Internet of things Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning
Divisions:	School of Computing > Master of Science in Cyber Security
Depositing User:	Ciara O'Brien
Date Deposited:	04 Jul 2026 14:10
Last Modified:	04 Jul 2026 14:10
URI:	https://norma.ncirl.ie/id/eprint/9475

Actions (login required)

View Item