on 09-04-2017 4:57 PM
I want to build a classification model in Predictive Analytics. As described in the documentation "Test (optional): used to compute the final model quality indicator." Can you explain for which specific indicators the test dataset is used? As I understand, test dataset does not influence the Ki and KR. I'm right?
For some reason I could not access the documentation page you mention.
As per the documentation below, the Test subset allows the user to Verify the performance of the selected model on a new dataset.
Note that the selection of the best model is always based on the validation subset whether you are using a Test dataset or not.
Here are examples of model performance graphs or reports where one can see the Test subset:
Display > Model Graphs > Datasets > All Datasets
Display > Statistical Reports > Model Performance > Estimator Variables / KI & KR / Other Performance Indicators.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
User | Count |
---|---|
80 | |
24 | |
11 | |
9 | |
7 | |
6 | |
5 | |
5 | |
4 | |
4 |
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.