<![CDATA[Background: The prevalence of obesity has become a global issue affecting all countries. Physical activity and sedentary behavior are believed to be key factors contributing to obesity. Objective: This study aims to examine the relationship between physical activity and sedentary behavior with Body Mass Index (BMI) using machine learning algorithms. Method: A total of 280 students from various programs at Universitas Pendidikan Indonesia participated in this study (101 males and 179 females), aged between 17 and 23 years. Physical activity was measured using the Actigraph GT3X accelerometer. Seven machine learning algorithms—including Naïve Bayes, Support Vector Machine (SVM), local k-nearest neighbors (KNN), Classification via Regression (CVR), decision tree, random forest, and artificial neural network (ANN)—were applied to predict obesity risk. The RapidMiner software was used for testing. Results: Based on the variables of physical activity, sedentary behavior, and demographic factors, SVM demonstrated the highest accuracy (74.22%) among the algorithms. For sensitivity and specificity, ANN and decision tree performed best, with values of 72.27% and 77.5%, respectively. Conclusion: Physical activity, total Metabolic Equivalent of Task (MET), and sedentary duration are significant predictors of obesity risk. Promoting physical activity and implementing campus policies are essential to reduce obesity prevalence among students.]]>
Copyrights © 2024
