Create a Decision Tree

Here are the steps to create a decision tree:

From the toolbar, click to create a decision tree.
From the Data pane, drag and drop the Age at Death variable into the Response field in the right pane.
In the Data pane, select Diastolic, Weight, Height, Cholesterol, Age CHD Diagnosed, Sex, and Cause of Death. Drag and drop these items into the model pane. The decision tree automatically updates.
Click in the upper right of the model pane.

In the details table, select the Node Rules tab. Notice that the only predictors used were Age CHD Diagnosed and Cause of Death. You can adjust the decision tree properties to include more predictors in the model.
Click the Properties tab in the right pane. The most obvious property to change is Reuse predictors. When you deselect this property, each predictor variable is used in at most one split. However, assume that reusing predictors creates the best split in each node for this example. This might not always be the case for your data.
Set the value of Maximum levels to 10. The decision tree now has a maximum depth of 10 levels, instead of the default 6. On the Node Rules tab of the details table, every predictor is used at least once.
Set the value of Maximum branches to 4. This allows each non-leaf node to split into at most four new nodes.
To view the Tree Overview window, click in the upper right corner of the exploration workspace. In the Tree Overview window, click to fit the entire decision tree into the Tree Overview window.

Although each node is difficult to see, your decision tree should resemble the following:
In the Tree Overview window, right-click, and select Derive a Leaf ID Variable. The default name for this variable is Leaf ID (1).

In the New Calculated Item window, click OK. The Leaf ID (1) variable appears in the Data pane.
Save the project.

Last updated: January 8, 2019