Enhancing user authentication through the implementation of the ForestPA algorithm for smart healthcare systems

Zaidi, Nurul Syafiqah and Mubarak Ali, Al-Fahim and Ahmad, Firdaus and Abubakar, Adamu and AlDharhani, Ghassan Saleh and Ab Razak, Mohd Faizal (2025) Enhancing user authentication through the implementation of the ForestPA algorithm for smart healthcare systems. Mesopotamian journal of Cybersecurity, 5 (2). 577-589.. ISSN 2958-6542

	PDF - Published Version Download (4MB)
	PDF - Supplemental Material Download (149kB)

Abstract

The machine learning-based authentication model for smart healthcare systems represents a crucial step in addressing the needs of an ever-evolving healthcare industry. The need to protect sensitive patient data, ensure regulatory compliance, and reduce medical errors, especially in the context of telemedicine and remote monitoring, underscores the importance of such systems. Traditional authentication methods frequently lack sufficient security, resulting in potential breaches. Relying solely on usernames and passwords, without supplementary authentication measures, exposes systems to advanced security attacks. As it involves patients’ health and human lives, it is important to provide additional authentication, fast machine learning-based authentication models and high accuracy at the same time. This study involves five participants with devices and performs various finger-based interactions (raising, lowering, moving the finger, applying pressure, adjusting orientation, and utilizing multiple hikes) while completing reading and image comparison tasks across multiple sessions. Each experiment lasted between 25 and 50 minutes for one participant, with reading tasks typically taking 10--15 minutes and image comparison tasks requiring 3--4 minutes, all measured in milliseconds. All these activities are recorded as a dataset for model training. A model was trained via the forest penalizing attributes (ForestPA) algorithm, which can classify profiles into real or fake profiles on the basis of their behavioral patterns. The results revealed a 99.99% accuracy rate in identifying fake profiles and avoiding them by accessing medical data even though they were able to bypass the username and password

Item Type:	Article (Journal)
Uncontrolled Keywords:	ForestPA, Key Forest-based, Classifiers, Medical, Internet of Things, User profile authentication
Subjects:	Q Science > QA Mathematics > QA76 Computer software
Kulliyyahs/Centres/Divisions/Institutes (Can select more than one option. Press CONTROL button):	Kulliyyah of Information and Communication Technology > Department of Computer Science Kulliyyah of Information and Communication Technology > Department of Computer Science
Depositing User:	Dr Adamu Abubakar
Date Deposited:	13 Oct 2025 12:35
Last Modified:	13 Oct 2025 15:42
URI:	http://irep.iium.edu.my/id/eprint/123674

Actions (login required)

View Item