Machine learning algorithms for high performance modelling in health monitoring system based on 5GNetworks

• Autores: Collantes Inga, Zoila Mercedes, Meza Carhuancho, Juan Luis, Vertiz-Osores, Jacinto Joaquin, Cueva-Rios, Maria Alina, Fuster-Guillen, Doris, Ocaña-Fernández, Yolvi
• Formato: artículo
• Fecha de Publicación: 2022
• Institución: Universidad Tecnológica del Perú
• Repositorio: UTP-Institucional
• Lenguaje: español
• OAI Identifier: oai:repositorio.utp.edu.pe:20.500.12867/6895
• Enlace del recurso: https://hdl.handle.net/20.500.12867/6895
• Nivel de acceso: acceso abierto
• Materia: Internet of things; Machine learning; Health care; Artificial neural networks; https://purl.org/pe-repo/ocde/ford#1.02.00

The development of Internet of Things (IoT) applications for creating behavioural and physiological monitoring methods, such as an IoT-based student healthcare monitoring system, has been accelerated by advances in sensor technology. Today, there are an increasing number of students living alone whoare dispersed across large geographic areas, therefore it is important to monitor their health and function. This research propose novel technique in high performance modelling for health monitoring system by 5G network based machine learning analysis. Here the input is collected as EEG brain waves which are monitored and collected through 5G networks. This input EEG waves has been processed and obtained as fragments and noise removal is carried out. The processed EEG wave fragments has been extracted using K-adaptive reinforcement learning. this extracted features has been classified using naïve bayes gradient feed forward neural network. The performance analysis shows comparative analysis between proposed and existing technique in terms of accuracy, precision, recall, F-1 score, RMSE and MAP. Proposed technique attained accuracy of 95%, precision of 85%, recall of 79%, F-1 measure of 68%, RMSE of 52% and MAP of 66%.