My goal is to print a decision tree model that I have saved as an XML file. I generated a decision tree model using the CHAID option on a balanced sample dataset based on literature supporting this approach when dealing with strongly unbalanced outcomes. I have applied it to a larger dataset of independent cases. Using the scoring wizard the model is applied successfully and returns an observed to predicted accuracy of around 70% for each binary outcome.
My question is how can I see a printed tree when I apply the model to each new dataset? I do not see this is the decision tree documentation. My interest is to share the visual representation of the model decisions with an audience and since it was built on a balanced sample the percentages in each node will not be representative of the new dataset it is applied to. If I am misunderstanding the this I would appreciate any clarity.
Best
Peter Spangler
=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD
Locked
