SAS Miner - missing value dot issue

[farahbalkis]

Hello, I am experiencing this "." issue ever since I've changed my training validation and testing to 2,1,0 in thedecision tree node, how do I fix this?

[PeterChristie]

The missing means there are zero False Negative observations. You are using a small dataset. Usually when large datasets are processed in Enterprise miner you have False Negatives and False Positives so the missing indicator has a number instead.

If you click on the ROC chart in the Model Comparison results and then click the table icon at the very top of the results window, you will see a table showing all zero values for False Negatives for the validation dataset.

BTW don't forget to set the model selection table property to "Validate" in your model comparison node. You do not want to select a model based on the training partition results as shown in your screen capture.