What are Model definition terms?
- Model name: what you want to call this model
- Email for notification: which email to send notifications to.
- Notification Limit: When you want to get a notification email?
- K Group Size Limit: Cross-validation or no Cross-validation
- Threshold Value: Smallest node size allowed. Larger = better model that runs faster.
- Spread of Responses: When splitting a node if the spread of responses is greater than this value, the model will not split beyond these two new nodes.
- Model Contrast: Overall contrast achieved between predicted and not-predicted nodes. If contrast achieved is greater than this value then model will stop running.
- Anonymous Data (No UID): The data set does not have unique identifiers.
- Available Data: What data sets have been imported and can be used to build models.
- Assigned Data: What data sets have been selected to build a model (must be at least one)