Apply constraints and targets to a model 

Create a Criteria Set for a trained model, then define constraints for input features and targets for output features. Send a completed Criteria Set and its model to the engine for evaluation.

Create or edit a Criteria Set

When you first open Optimize, you will see the Optimize a model page. From here, you can create, edit or clone Criteria Sets to apply to trained models, and view the results of evaluating a model against a Criteria Set.

To create a new Criteria Set:

  1. Click Create new Criteria Set.
  2. In the dialog, enter a Name for the model, and Tags if required.
  3. Select a trained model from the drop-down list.
    • This should be the model you want to apply these criteria to.
    • The model cannot be changed once the Criteria Set has been created.
  4. Click Save.

To clone an existing Criteria Set: 

  1. Click a Criteria Set in the list to select it, then click Clone Criteria Set or right-click and select Clone.
  2. In the dialog, edit the Name and Tags.
  3. Click Save.

Define or edit constraints and targets

After saving a new Criteria Set, the criteria definition page is displayed. Input features are displayed on the left hand side, and output features on the right. Changes on the criteria definition page are saved automatically.

You can come back to this page to edit an unoptimized Criteria Set by right-clicking it on the Optimize a model page and clicking Open.

Set constraint type and value for inputs

By default, all model input features are included in the Criteria Set unconstrained.

To apply a constraint to the model:

  1. Select a constraint type from the drop-down list in the Constraint column (Full Range, Fixed or Between).
  2. For Fixed or Between constraints, edit the allowed Value for the input feature.

Set goal type and target value for outputs

By default, output features are not assigned a target value and are equally weighted.

To set a target value for an optimization problem:

  1. Set a goal type from the drop-down list in the Goal column (Greater than/Above, Less than/Below, Between or None).
  2. Enter a target Value for the output.
    • The Value field auto-populates with the maximum or minimum of the model Range.
    • Click the Value field after selecting a goal type of None to clear it.
  3. Edit the Weight (importance) of the output relative to the others.

Edit Criteria Set name or tags

To rename a Criteria Set or edit its tags:

  1. Open the Edit Criteria Set dialog:
    • On the Optimize a model page, select the Criteria Set in the list and click Edit Criteria Set, or right-click and select Edit Criteria Set .
    • On the criteria definition page, click the Edit Criteria Set icon next to the Criteria Set name. 
  2. In the dialog, edit the Name and Tags.
  3. Click Save.
Tip: The model cannot be changed once the Criteria Set has been created. You must create a new Criteria Set to use a different model.

Run an optimization problem

To send a Criteria Set and its model to the engine, click Optimize on the criteria definition page.

The card for your Optimization job will be displayed here and in the MI Job Queue.

Edit and re-run an optimized Criteria Set

Criteria Sets cannot be edited or re-sent to the engine once they have been optimized; you must Clone the Criteria Set and edit the cloned Set.