IoT guardian: explainable deep ensemble learning for reliable security in Internet of Medical Things

When it comes to smart healthcare business systems, network-based intrusion detection systems are crucial for protecting the system and its networks from malicious network assaults. To identify network-based intrusions in Internet of Medical Things (IoMT), this paper lays out an ensemble method based on deep learning that makes use of patient biometrics and characteristics of network traffic. Medical features that are complicated, non-linear, and overlapping may be learned using random forest feature significance. By feeding meta-learner (MLP) predictions from weak learners (CNN and LSTM), an enhanced deep-stacked ensemble method with explainable decisions is achieved. The suggested model demonstrated a detection accuracy of 96% on the set of networks and patient biometric characteristics that were pre-selected, and 94% on the set that were not pre-selected. Experiments conducted on various state of the art deep learning methodologies illustrate the resilience and adaptability of the proposed model. The suggested technique demonstrated superior performance compared to the current approaches across all test situations, with a notable improvement in accuracy of 1-3% on the IoMT intrusion dataset. To protect IoMT devices in addition networks from intruders in healthcare in addition medical settings, the suggested model may be used as a tool for monitoring IoMT networks.