Decision Tree

Select Decision Tree on the Model screen to start the creation process.

1. Classifier (drop-down list): Choose a field from the index to act as a classifier. All fields from the currently loaded index are available. The classifier in a decision tree is the dependent variable used to discover the relationships with other variables.

2. Input Variables: To change the number of input variables that the model will be built from, use the left and right arrows to move fields between the Unselected and Selected boxes. Fields in the Selected box will be included in the model.

   Note: Fields with numerical data should be banded before running the model to minimise the impact of individual anomalies. Fields with a large amount of unique values will not be included in the model.

3. Options: Enable Use AdaBoost option to enable adaptive boosting and enable Attribute Selection: to simplify the output. Please refer to Key Terminology for more detailed information.

Click Create Decision Tree to start the generation process. Once finished a Decision Tree generated network is produced.

To set up a Bayesian Network click the icon on the left of the screen.