Fig. 3
From: Machine learning in physical activity, sedentary, and sleep behavior research
Three commonly employed machine learning algorithms with hypothetical data for predicting active and inactive individuals (a) A decision tree model comprising two levels. In decision trees, the selection of the appropriate attribute for each node at every level, the correct cut-off point, and the identification of relevant categories is typically accomplished using metrics such as Information Gain and Gini Index. b Random forest model comprised of three decision trees. c SVM model for a binary classification problem with two features