This menu can be used to form a regression tree, which is a mechanism for predicting a response variable from a set
of independent variables. The construction process splits the observations into subsets, according to whether or not
they are less than a particular value of one of the independent variates. The aim is to form subsets that have similar
values for the response variate. The predicted value of the response variable for each node of the tree is the mean of
its value for the subset of observations at that node. The accuracy of the node is the squared distance of the values
of the response variate from their mean for the observations at the node, divided by the total number of observations.
The potential splits at the node are assessed by their effect on the accuracy, that is the difference between the accuracy
of the node and the sum of the accuracies of the two potential successor nodes. The node will become a terminal node if
none of the splits provides any improvement in accuracy, or if the mean square of the observations at the node is less
than a specified limit.
Available Data
This lists data structures appropriate to the current input field. The contents will change as you move from one field to
the next. Double-click on a name to copy it to the current input field; alternatively, you can type the name directly into
the input field.
Y-Variate
Provides a space for you to specify a response variate for the regression.
X-Variables
Specifies the independent (x) variables available for constructing the tree. The 
button allows multiple selections to be copied from Available Data.
Save Tree in:
Specifies an identifier name to save the resulting tree in. The tree will be saved within a GenStat Tree data structure.
Prune
The construction of a regression tree generally results in over fitting, that is it continues to extend the branches of
the tree beyond the point that can be justified statistically. One solution is to prune the tree to remove the uninformative
sub-branches. Clicking this button opens a menu where you can prune the tree.
See Also