Machine learning models for data analysis and protection in IoT-based healthcare administration systems

DOI: 10.31673/2412-9070.2026.318110

Abstract

The article examines the application of machine learning models for the analysis and protection of data within automated medical administration systems operating on the basis of Internet of Things technologies. Modern healthcare institutions increasingly integrate IoT devices for patient monitoring, equipment control, environmental data collection, and optimization of administrative processes. As a result, the volume, heterogeneity, and dynamic nature of medical and infrastructural data continue to grow, which complicates their processing and significantly increases the number of potential vulnerabilities. At the same time, IoT-enabled infrastructures introduce additional risks associated with cybersecurity threats, unstable communication channels, device malfunctions, and various ano-malies in data streams that may negatively affect the reliability of medical services.
This work analyzes the capabilities of Decision Tree and Random Forest algorithms in adderssing classification, prediction, and anomaly detection tasks in medical information systems. Both methods demonstrate high interpretability and accuracy, which is crucial for decision-making processes in the healthcare domain. A conceptual architecture for integrating machine learning models into a multi-level medical administration system is proposed, detailing the roles of data acquisition, communication, analytical, and administrative layers. The implementation of machine learning modules enhances data quality assessment, strengthens system security through behavioral analysis of IoT traffic, and supports intelligent decision-making for resource planning and operational management in medical centers.
The presented examples demonstrate the practical potential of ML-based solutions for improving the reliability, stability, and security of IoT-oriented healthcare infrastructures. The study confirms that the combined use of IoT and machine learning technologies significantly increases the efficiency of medical administration processes and opens new opportunities for the development of intelligent clinical and organizational support systems.

Keywords: IoT, machine learning, medicine, data analysis, cybersecurity, automation, Decision Tree, Random Forest.