A set of tools is included for training and pruning a classification tree. The monitoring graph
shows the incremental inclusion of the training pixels per class during the analysis. Once the
analysis is complete, the proportion of misclassified training site pixels will be displayed as well
as a tree showing the splits for the relevant input images and the split threshold. Leaf details
include the number of pixels in a leaf, the percentage of these pixels that belong to the total
pixels in that class, and a purity index that is the percentage of correct pixels belonging to the
specified class in that leaf to the total number of pixels in the leaf.