i am working with the view supplied with the db from the Spofford book on learning MDX to try to learn about decision trees because i couldnt figure out how to see the very beginnings of the model from Foodmart. My problem is that the decision tree only has one node.
In searching BOL, i found two paramters called COMPLEXITY PENALTY and MINIMUM LEAF CASES that i thought would help me see the discrimination in the data better, but varying these two params by hi and low against each other did not produce a tree, just one node.
Can i buy a clue as to what i am doing wrong to mask the branching?
I called Brand Name the case and called it keys unique, Promo Name is an input only, SKU Name is predictable, but all i get is one lonely node.
thanks very much for your help
drew