Data Science Desktop Survival Guide
by Graham Williams
Rattle Model Tab
We can click directly on the Model tab, skipping over the Explore, Test, Transform, Cluster and Associate tabs, to display the modelling options. The default modelling option is to build a decision tree, and it's always a good place to start.
Various options to tune the building of a decision tree are provided in the tab.
Rattle deploys rpart by default to build the decision tree, and many more options thatn are presented through the rattle GUI are available through using rpart() directly. These will be explored later in this chapter.