How to change a threshold to predict Good or Bad in the Enterprise Miner

[shofuku]

We are developing some credit scoring models using Credit Scoring for Enterprise Miner. We can predict Good(non-event) or Bad(event) by developed model in SCORE node. It's threshold probability to identify Good or Bad is set to 0.5. But our threshold to identify Good is higher than 0.5.

Could anyone tell me how to change this threshold?

Thanks in advance.

[WendyCzika]

You can use the Cutoff node (from the Assess tab), then set the Cutoff User Input to whatever value you want to use for the cutoff.  The variable EM_CUTOFF will then have the predicted level based on this cutoff.

[shofuku]

Hi Wendy,

Thank you for your prompt reply.
I could change a threshold value by using Cut off node and SAS Code node
to insert them before SCORE node.

I have one more question about threshold.
Can I change a threshold to calculate misclassification rate in Score
Card node or other modeling modes such as Tree, HP Forest?