Boosted Trees Results
Click the
OK button in the
Boosted Trees Specification
You can also request to perform additional boosting steps (for example, request More trees) or select a specific number of boosting steps (Number of trees) as the final results. In this case, all predicted values or classifications, as well as deployment code (for scoring or predicting new cases, in predictive data mining applications), will be computed based on the requested number of trees.
The tabs available in this dialog depend on the particular analysis in question, i.e., whether the current analysis is a
Regression Analysis or
Classification Analysis, as selected in the
Boosted Trees Quick
The options described here are available in the Boosting Trees Results dialog regardless of the tab selected.
More trees/Number of more trees
Enter a number in the
Number of more trees box, and click the
More trees button to perform an additional number of boosting steps, such as to compute more trees (see the
I ntroductory Overview
Number of trees
Specify the number of boosting steps that you want to retain as the final model.
By default, the program will determine the best final model as the one (for instance, as the specific number of trees) that yielded the smallest error estimate for the testing sample. You can use the
Summary option on the
Boosted Trees Results Quick lassification
Cancel
Click the
Cancel button to close the dialog, and return to the
Boosted Trees Specifications
Options
Click the Options button to display the Options menu.
By Group
Click the By Group button to display the By Group specification dialog.