The Reliability of Classification of Terminal Nodes in GUIDE Decision Tree to Predict the Nonalcoholic Fatty Liver Disease

Figure 1

GUIDE classification tree with estimated priors probability (the proportion of patients in each class) and equal misclassification costs for predicting NAFLD. At each intermediate node, an observation goes to the left branch if and only if the condition is satisfied. Dark nodes represent predicted class “with NAFLD” and white nodes represent predicted class “without NAFLD.” Each terminal node has been formed in boxes of 3 parts so that the specified section in the left side of the box represents the number of individuals in the trained dataset who have been placed in this node. Specified percentages for this dataset are ratio of patients with NAFLD. The middle specified section of the box exhibits the node’s number and specified percent below it shows the overall ratio of the test datasets which have been placed in this node. The specified section in the right side of the box shows the number and percentage of the test datasets that have really high risk of NAFLD.