Research On Intelligent Recommendation Of Exercise Prescription For College Students’ Physical Health Based On Machine Learning

The physical health of college students has attracted more and more attention from all walks of life.As a new force in national development,college students’ physical health is related to whether the country has enough strength to meet the difficulties and obstacles on the road to prosperity.However,with the advent of modernization of science and technology,people’s lifestyles have undergone great changes,and college students’ physical health is also declining year by year.Machine learning methods are being applied to all aspects of social life in various forms,providing convenience for people’s lives in many areas.This article mainly aims at the physical health of college students,using machine learning methods,proposes to recommendation and self iteration of exercise prescription for college students with different constitutions.The main work of this thesis is as follows:(1)This thesis proposes a classification of College Students’ physical health based on k-medoids method.We conducted extensive exchanges and in-depth investigations on the physical health of college students with teachers and experts from Chengdu Physical Education Institute.Students with different physical health should have different training items,training intensity and methods in the exercise prescription.In order to more accurately recommend exercise prescription for college students with different physical health status,after the research and experiment of various classification algorithms,it is found that k-medoids clustering algorithm has higher classification accuracy when classifying college students’ physical health data.(2)This thesis designs a prediction method of College Students’ exercise prescription based on convolutional neural network.Based on the classification of College Students’ physical health,this paper studies the prediction effect of exercise prescription under various neural networks.Comparative experiments were conducted on the current mainstream prediction models,including MLP,BPNN,SVM and CNN,and finally the Convolutional neural network of best effect was adopted to realize the prediction and recommendation of exercise prescriptions for college students of different physical fitness.(3)This thesis designs a self-adjusting method of exercise prescription based on NLP emotion analysis.After the initial exercise prescription is recommended for college students with different physical health problems,the heart rate data and NLP emotion analysis data are further combined to self iterate the corresponding exercise prescription,so as to achieve more targeted personalized exercise prescription recommendation.(4)In this thesis,a prototype system of intelligent recommendation of College Students’ physical health exercise prescription based on machine learning is implemented.The theoretical knowledge combined with the existing methods is fully applied to the actual system,mainly including the system preliminary prediction of exercise prescription,self-adjustment of exercise prescription according to the feedback data after exercise and other modules,and realized intelligent recommendation of College Students’ physical health exercise prescriptionOther works of this paper includes: creating Exercise Moods exercise emotion data set and designing 20 exercise prescription templates.