Task 2: Uploading a PMML file and connecting a predictive model

You upload a PMML file to create a machine learning model deployment and connect a predictive model to the machine learning model.

Step 1: Exploring the PMML model

Open the Standard scaler PMML file and browse through it:

  • The input schema, which consists of 7 fields, is defined in the DataDictionary section.
  • The output schema, which consists of 3 fields, is described in the Output section. This section is optional.
  • It uses a regression model.

Step 2: Uploading a PMML file

  1. From Task 1, you should be in the Configure predictive model wizard.
  2. Click Select provider and deployment and choose the embedded-quick provider.
  3. Click Upload and add the Standard scaler PMML file.
  4. Click Upload again. A new machine learning model deployment is created and selected. You are redirected to the next step to define the input schema.
  5. Review the input schema: it is already complete, just browse through it.
  6. Click Next to check the test invocation.

    This step is used to invoke the model and generate an output example, which can be used to define the output schema. However, since the output schema is already specified in the PMML file, you can skip this step.

  7. Click Next to review the output schema.

    The output schema, as defined in the PMML file is already complete. Just browse through it.

  8. Click Apply to connect the predictive model.

    The predictive model diagram opens, now configured with an embedded provider and the model deployment you just loaded.