A Model for Predicting Physical Health of College Students Based on Semantic Web and Deep Learning Under Cloud Edge Collaborative Architecture

Abstract

A model for predicting physical health of college students based on semantic web and deep learning under cloud edge collaborative architecture is proposed to address the issue of most physical health prediction models being unable to fully describe the characteristics of sports performance changes and having large prediction errors. Firstly, the authors design a measurement data analysis system based on cloud edge collaboration architecture to improve data analysis efficiency. Then, they preprocess the data on the edge side, such as missing samples, and extract data features using an equal dimensional dynamic GOM model. Finally, they deploy the RBFNN-SSA model in the cloud center, input the characteristics of each indicator into the model for predictive analysis, and obtain the physical health status of college students. Based on the physical health test data of Hohai University from 2018 to 2021, an experimental analysis was conducted. The results showed that all three intervention measures had significant effects on maintaining and improving the physical health level of college students.