Explainable Ensemble-Based Detection of Cyber Attacks on Internet of Medical Things

As new applications of the Internet of Things emerge in the health and medical services, malicious actors target these applications increasingly. These growing application bring a variety of privacy and security challenges. In this paper, we present an explainable machine learning ensemble designed to detect attacks on Internet-of-Medical-Things with high accuracy. The proposed system was tested using WUSTL-EHMS-2020 dataset. Tests showed that the proposed ensemble is capable of delivering an accuracy exceeding 99%, with an score exceeding 0.99. The proposed system was explained using SHAP values to provide insights into the most impactful features, and the nature of their impact on the system's decisions.