Defining Model Properties

  1. Under Primary Stages / Deployed Stages / Analytics Scoring, click the Java Model Scoring stage and drag it onto the canvas, placing it where you want on the dataflow and connecting it to input and output stages. Note that the input stage must be the data source that contains both the objective and input variable fields for your model. If you are running your job in batch mode, you will also need an output stage to capture model scores; otherwise you will use a Spectrum™ Technology Platform web service to score data in real time.
  2. Double-click the Java Model Scoring stage to show the Java Model Scoring Options dialog box.
  3. Optional: Select the type of a model you are scoring in the Type filter drop-down.
  4. Select the Type filter being used to score the model.
  5. Select the Model name from the drop-down.
  6. Enter the type of model you are scoring in the Model type field.
  7. Optional: Enter a Description of the model.
  8. Check Ignore unknown categorical levels to return data in the Predicted_Value column for rows with a categorical level that was not in the data used to fit the model. If you leave this box unchecked, the Predicted_Value column will return "Null/NA" for those rows.
  9. The Inputs table shows information for the model's input fields. These fields and their data types automatically map to Spectrum fields and data types.
  10. Click OK to save these options or continue to the next tab.