The most accurate tree is the tree with the lowest misclassification cost. This tree is also known as the optimal tree.

Sometimes, simpler trees with slightly higher misclassification costs work just as well. Typically, a tree with fewer terminal nodes gives a clearer picture of how each predictor variable affects the response values. A smaller tree also makes it easier to identify a few target groups for further studies. If the difference in prediction accuracy for a smaller tree is negligible, you can use the smaller tree to evaluate the relationships between the response and the predictor variables.

Click Select an Alternative Tree to open an interactive view of the plot that includes a table of model summary statistics. Use the plot to investigate smaller trees with similar performance.

In this example, the tree with 4 terminal nodes has the label "Optimal"
because the criterion for the creation of the tree is the smallest tree with a
misclassification cost within 1 standard error of the minimum misclassification
cost. The tree with 4 terminal nodes has a misclassification cost of
approximately 0.415. The tree with 6 terminal nodes has a slightly lower
misclassification cost of approximately 0.397. The tree with 7 terminal nodes
has the minimum misclassification cost of approximately 0.391. The initial tree
with 4 terminal nodes keeps the "Optimal" label when you use
Select
an Alternative Tree
to create results for a different tree.