Task 1: Defining a machine learning provider in Decision Designer

You import and explore a new decision service. Then, you define a machine learning provider using Watson Machine Learning.

Step 1: Importing a sample decision service

You import a sample decision service into your decision automation. This decision service applies several criteria to determine a borrower’s loan eligibility. One key factor is risk, which is predicted by the machine learning model.

  1. Sign in to Decision Intelligence using your instance credentials.
  2. Create a decision automation.
  3. Click New decision +.
  4. Click Discovery tutorials and select Machine learning quick tutorial - Loan approval. Then, click Import.
  5. Open the Quick machine learning loan approval decision service.
  6. Explore the imported decision service:
    1. Open the Loan Validation Data data model in the Data tab to review the defined types. Note that the risk probabilities type is used by the predictive model.
    2. Navigate back to the decision service name in the breadcrumbs and open the Loan Validation Decision Model. This model determines whether a loan can be approved. One key decision is the Risk Score node, prediction node linked to the loan risk score predictive model.
    3. Navigate back to the decision service name in the breadcrumbs and open the loan risk score predictive model. It is not connected yet. You will connect it in the next steps.
    4. Review the input and output mapping nodes:
      • Click the Input mapping node to see how the machine learning model input is derived from loan and borrower data.
      • Click the Output mapping node to see how the output of the machine learning model is converted into a risk probability.

Step 2: Defining a machine learning provider

You configure a machine learning provider to use your deployed model in the decision service. For more information about machine learning providers, see Managing local machine learning providers.

  1. Return to the diagram and click anywhere outside the nodes or links.
  2. Click Configure.

    The configuration panel opens, showing that no provider is defined.

  3. By default, Remote Machine Learning Model is selected. Click Next to continue.
  4. Click New + to create a provider:
    1. Type: Select Embedded machine learning.
    2. Name: Enter embedded-quick.
    3. Description: Enter Provider for the machine learning quick tutorial.
  5. Click Test connection. It successfully connects to the provider.
  6. Click Save. Wait until the Status shows Running to confirm the provider is ready.
  7. In the Navigation history, select loan risk score to open the predictive model.