Classifying High School Students’ Health-Related Physical Fitness Report Cards With Data Mining

Abstract

In this study, classifications were made from the data obtained from the Health-Related Physical Fitness Report cards and BMIs of students through data mining methods, artificial neural networks, and decision trees models. Then the classification performances of both models were compared. The body weight and height measurements of the students in the Health-Related Physical Fitness Report were formulated, and their BMI classification was made. In addition, it was investigated whether other parameters such as shuttle, push-up, and sit-and-stretch flexibility test values had an effect on BMI classification. The study group comprised 1,050 secondary school students studying in the Cihanbeyli district of Konya in 2017. In conclusion, it was determined that artificial neural networks had more correct classifications than decision trees analysis. In the Health-Related Physical Fitness Report, shuttle and push-up stood out among the variables affecting BMI classification.