Machine Learning-Based Detection of Internet of Thing Attacks in Healthcare Environments
Li-ChuWu, Chia-Mei Chen, Zheng-Xun Cai, Ming Hsia Hsu, Wang‐Chuan Juang · 2021
The emerging technologies of the Internet of Things (IoTs) have started a new chapter for healthcare. IoT technologies have transformed the way healthcare delivering care to patients. As the network-enabled IoT devices are integrated with healthcare network infrastructure, IoT security is a major concern for medical institutes. This study proposes a machine learning based detection model to identify malicious behaviors in such IoT network environments. The proposed two-phase LSTM detection method first identifies the network protocol of the traffic and then detects IoT anomalies. As most data is imbalanced with a small portion of malicious traffic, the study demonstrates the impact of imbalanced data in model training and suggests an effective approach to handle such a situation. The experimental results show that the proposed two-phase LSTM classification model outperforms onephase one and other classification models