Title: Implementation of fitness and health management system utilising deep learning neural network and internet of things technology
Authors: Xiaojun Zhang
Addresses: Xi'an Polytechnic University, Xi'an, Shaanxi, China
Abstract: The purpose is to implement fitness and health management services more scientifically, enhance people's awareness of health management, prevent diseases caused by long-term sub-health, and comprehensively improve people's fitness and health status physically and mentally. Specifically, the data of people's health indicators are analysed, and a fitness and health management service system is established using deep learning and Internet of Things (IoT) technologies. First, people's fitness and health indicators are detected using IoT technology and integrated and pre-classified into text, number, and image. Afterward, the pre-classified data are input into the Convolutional Neural Network (CNN), their features are extracted for modelling and analysis, and the results are input into the constructed BP BackPropagation Neural Network (BPNN) model. Consequently, a preliminary prediction result about the user's fitness and health is obtained for the user's fitness and health status. The results show that the constructed fitness and health management system based on the proposed ensemble prediction model is more optimised than those constructed by a traditional simple model. With the proposed intelligent fitness and health management system composed of IoT devices, users can gain a better health status by self-monitoring, self-control, self-discovery, self-analysis and self-search.
Keywords: health management; backpropagation neural network; DS evidence theory; composite prediction model.
International Journal of Grid and Utility Computing, 2022 Vol.13 No.1, pp.66 - 75
Received: 18 Feb 2021
Accepted: 04 Jul 2021
Published online: 11 Mar 2022 *