Research And Application Of Exercise Prescription Generation Technology Based On Collaborative Filtering Model

Health is people's basic demand for a better life,and more and more people keep healthy through exercise.Technical research on the generation and management of exercise prescriptions plays an important role in promoting people's high-quality fitness exercises.Therefore,the application of exercise prescription has received great attention.But,aiming at the problems of insufficient personalization,poor pertinence,and unreasonable exercise planning in the existing exercise prescription generation methods,this thesis combines the relevant theories of exercise of exercise prescription to research and design a personalized exercise prescription generation method.On that basis,in the view of the safety issues in the implementation of exercise prescriptions,an exercise risk monitoring model based on capsule neural network is constructed to reduce the risk during exercise.We verified the reliability of related technologies by establishing an exercise prescription service system.The specific research content is as follows.First of all,this thesis quantifies the functional attributes of sports,based on the relationship between fitness needs and sports.We calculate the fitness of sports to sports needs based on the quantified value of the functional attributes of sports.On this basis,we built the generation method and management method of exercise prescription.On the one hand,we integrated the focus coefficient matrix into the collaborative filtering calculation,so that,we can use it to calculate the fitness for specific sports need,and then,use this as a basis to select sports that meet expectations.On the other hand,we use the non-dominated sorting genetic algorithm to calculate the parameters of exercise prescription such as exercise time,exercise intensity,exercise frequency to ensure the pertinence and rationality of the exercise prescription.According to the selected exercise items and calculated exercise parameters,we can complete the generation of exercise prescription.The experimental results show that the collaborative filtering algorithm fused with the focus coefficient matrix has better accuracy and sparsity in the selection process of sports items.Secondly,this thesis constructed an exercise risk monitoring model based on capsule neural networks to ensure the safety of exercising,because of that,the heart rhythm data from wearable device can reflect the body state during exercise in the round.This model extract features of heart rhythm data collected by wearable devices based on the vector neurons.And then,we use the capsule neural network to estimate the risk of exercise from the features of heart rhythm data,so as to avoid the occurrence of dangerous situations during exercise.In addition,the capsule neural network based on vector neurons can extract the correlation information between features.So that,we can make the model of exercise risk resistant to the noise interference caused by movement in the data.And then,we created a deep capsule neural network to monitor abnormal heart rhythms during exercise.From the experimental results,we can get that the exercise risk monitoring model based on the capsule neural network has high accuracy and reliability in the detection of arrhythmia under the noise interference environment.The deep capsule neural network also has faster performance and prediction accuracy than convolutional neural network.Finally,this thesis designs an exercise prescription management process based on feedback adjustment mechanism to ensure the effectiveness of exercise prescription implementation.This mechanism can adjust sports,and optimize the motion parameters of the exercise prescription,with the feedback of preference and physical fitness of users during the implementation process of exercise prescription.This mechanism makes up for the lack if analysis of user differences in the process of generating exercise prescriptions,and ensures the rationality and efficiency of exercise prescriptions.Exercise prescription management is really the guarantee of exercise effects,it makes the exercise prescription in this thesis more adaptable in the application process.On this basis,this thesis designs and implements a personalized exercise prescription service system,which verifies the feasibility and practicability of the related methods of exercise prescription generation and management in this thesis.