Font Size: a A A

Exercise Persistence Studies Based On Classification And Prediction Of Exercise Behavior

Posted on:2017-11-05Degree:MasterType:Thesis
Country:ChinaCandidate:Y LiuFull Text:PDF
GTID:2417330623454764Subject:Management Science and Engineering
Abstract/Summary:PDF Full Text Request
National fitness becomes the key project of sports industry.With people's growing enthusiasm of public fitness,the national physical quality needs to be improved.In order to keep people motivated and ensure effectiveness and persistence of physical exercise,this dissertation makes a series of studies on how to scientifically and effectively evaluate and predict the exercise behavior to improve the physical activity level and exercise persistence.In this dissertation walking is the representative exercise mode of public physical exercises.According to the domestic and foreign literature,three kinds of physical activity levels and standards that are suitable for the general public are established.Based on walking steps generated by individual,features such as stride frequency and duration of exercise are extracted.Then we research on exercise behavior classification using support vector machine with different types of kernel functions,and the features are extracted from walking steps generated by different people with different age,sex and BMI index.And different Markov model prediction methods are proposed,such as the first order Markov chain to predict individual exercise behavior and evaluate its effect and persistence.The exercise behavior classification model with high accuracy is established by adjusting the extracted features and classier parameters.At the same time,transition probability matrix is established according to the individual exercise state sequence,finally the exercise behavior prediction model based on Markov chain is established with high prediction accuracy.By the establishment of the model,we can classify and identify the data generated by individual accurately and predict the future exercise state.According to the stationary distribution we can also evaluate and predict the individual exercise persistence.The study of this paper is of great significance for the evaluation and prediction of the mass exercise behavior,and it provides the basis and ideas for further research,such as the application of exercise prescription to enhance the mass exercise persistence.
Keywords/Search Tags:exercise behavior, exercise persistence, walk, physical activity level, support vector machine, Markov chain
PDF Full Text Request
Related items