Tsetlin Machine-based Intrusion Detection System for IoMT Networks

Forum topic · 小凯 · 2026-04-06

Summary

This paper proposes a novel Tsetlin Machine (TM)-based intrusion detection system (IDS) for protecting Internet of Medical Things (IoMT) networks from cyberattacks. The TM is a rule-based, interpretable machine learning approach that models attack patterns using propositional logic, offering transparency that is critical for healthcare security. The rapid adoption of IoMT enables seamless connectivity among medical devices and services but introduces serious cybersecurity and patient safety risks, as attackers exploit emerging vulnerabilities to infiltrate these networks. Evaluated on the CICIoMT-2024 dataset, which covers multiple IoMT protocols and cyberattack types, the proposed TM-based IDS achieves 99.5% accuracy in binary classification and 90.7% in multiclass classification, outperforming existing state-of-the-art methods. The work is authored by Rahul Jaiswal, Per-Arne Andersen, and Linga Reddy Cenkeramaddi, and was released on arXiv (2604.03205).

Overview

Field: Machine Learning Authors: Rahul Jaiswal, Per-Arne Andersen, Linga Reddy Cenkeramaddi, et al. Published: 2026-04-03 arXiv: 2604.03205

Abstract (Original)

The rapid adoption of the Internet of Medical Things (IoMT) is transforming healthcare by enabling seamless connectivity among medical devices, systems, and services. However, it also introduces serious cybersecurity and patient safety concerns as attackers increasingly exploit new methods and emerging vulnerabilities to infiltrate IoMT networks. This paper proposes a novel Tsetlin Machine (TM)-based Intrusion Detection System (IDS) for detecting a wide range of cyberattacks targeting IoMT networks. The TM is a rule-based and interpretable machine learning (ML) approach that models attack patterns using propositional logic. Extensive experiments conducted on the CICIoMT-2024 dataset, which includes multiple IoMT protocols and cyberattack types, demonstrate that the proposed TM-based IDS outperforms existing state-of-the-art methods.

Key Results

Binary classification accuracy: 99.5%
Multiclass classification accuracy: 90.7%
Dataset: CICIoMT-2024, covering multiple IoMT protocols and cyberattack types
Approach: Tsetlin Machine — rule-based, interpretable ML using propositional logic to model attack patterns

The results suggest that interpretable, logic-driven learning can serve as an effective alternative to deep learning-based IDS approaches in safety-critical healthcare network environments.

--- *Auto-collected on 2026-04-06*