Decision and Recommendation System Services for Patients Using Artificial Intelligence
Arjun Tandon, Raghav Dangey, Anand Kumar Mishra, G. Balamurugan, Amit Kumar Tyagi · 2023
Recent advancements in the Internet of Things (IoT) have made it possible to create intelligent surroundings. With any technology, the IoT idea is thought to carry considerable security and privacy threats. The various potential incursions that may be made pose privacy and security issues. The creation of an intrusion detection system is necessary to identify attacks and anomalies in the IoT system. This study presents a Deep Belief Network (DBN) algorithm model for the intrusion detection system. The CICIDS 2017 dataset is used in the performance analysis of the current IDS model for assaults and anomaly detection. All of the metrics, including accuracy, recall, precision, F1-score, detection rate, and others, were improved by the recommended technique. The way that people get healthcare has been transformed by IoT technology. Due to the integration of network-enabled IoT devices with healthcare network organizations, medical institutions are extremely worried about IoT security. This research presents a machine learning-based detection mechanism for dangerous behaviors in such IoT network contexts. The suggested two-phase LSTM detection technique establishes the network's traffic protocol to detect IoT abnormalities. The study illustrates the consequences of unbalanced data on model training and provides a workable solution because the bulk of the data is unbalanced and there is just a small quantity of malicious traffic. The experimental results show that the proposed two-phase LSTM classification model outperforms other classification models, as well as one-phase one.