Khyoi Nu, Tahar Touati, Srushti Buddhadev, R. Sun, M. Smuck, I. H. J. Song
{"title":"Who is physically active? Classification and Analysis of Physical Activity using NHANES data","authors":"Khyoi Nu, Tahar Touati, Srushti Buddhadev, R. Sun, M. Smuck, I. H. J. Song","doi":"10.1109/SSCI47803.2020.9308353","DOIUrl":null,"url":null,"abstract":"Physical activity (PA) brings health benefits to adults. It is a crucial indicator of the general health condition, whether a person is physically active or not. This paper proposes ML (Machine Learning) -based PA classifiers to predict the individual PA level for each person. Besides, the proposed classifiers extract the determinants that identify an active person. The classifiers yield an AUC of up to 0.81 and specificity and sensitivity of up to 0.79. From the classifiers, we conclude that age and gender are the most influential determinants. Notably, body mass index (BMI) impacts females more strongly than males, whereas screen time for TV impacts males more strongly. The result of the study guides a proper type of PA intervention and provides an efficient way to engage in personalized health programs and medical treatments.","PeriodicalId":413489,"journal":{"name":"2020 IEEE Symposium Series on Computational Intelligence (SSCI)","volume":"06 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2020-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"1","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2020 IEEE Symposium Series on Computational Intelligence (SSCI)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/SSCI47803.2020.9308353","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 1
Abstract
Physical activity (PA) brings health benefits to adults. It is a crucial indicator of the general health condition, whether a person is physically active or not. This paper proposes ML (Machine Learning) -based PA classifiers to predict the individual PA level for each person. Besides, the proposed classifiers extract the determinants that identify an active person. The classifiers yield an AUC of up to 0.81 and specificity and sensitivity of up to 0.79. From the classifiers, we conclude that age and gender are the most influential determinants. Notably, body mass index (BMI) impacts females more strongly than males, whereas screen time for TV impacts males more strongly. The result of the study guides a proper type of PA intervention and provides an efficient way to engage in personalized health programs and medical treatments.