| Rank Fields | Induce Single Node | |
The Tree tab of a MinedTree object is designed for developing and testing decision (classification) trees from data defined in the Training Table on the Tables tab 
using the attributes and outcome definitions of the Fields tab 
.
The Tree tab shows the profiling of data in terms of tree branching according to attributes leading to terminal decision points (leafs) in this typical (default) format. In the case of discrete outcomes, each leaf on the tree shows an outcome followed by two figures, the first figure represents the number of records from the data sample which matches that tree profile, while the second figure represents the probability of these records having the same outcome as the leaf. In the case of continuous outcomes, each leaf in the tree shows the average value and the standard deviation, together with the total frequency of records in that leaf.
The tree offers the following features:
·Induction of graphical trees.
·Interactive induction of trees.
·Building of trees from discrete or continuous outcomes.
·Verification of the tree against independent data. A specified percentage of the data sample is automatically set aside for testing.
Note that before you can induce a decision tree you must set the outcome using the Fields tab.