Let's now step through some initial illustrations of using the model builder:

1. To use the model builder, you need to add the Watson Machine Learning service as an asset to a new or existing project (this is referred to as associating the service with the project). To do this from within your project, you can click on Add to project and then choose the WATSON MACHINE LEARNING service as the asset type, as shown in the following screenshot:

2. Once you have added this service to your project, you can then create and add new models to the project simply by clicking on New Watson Machine Learning model under Models:

3. On the New model page (as shown in the following screenshot), you must provide your new model's basic details as follows:

• Name: The name of your new model.
• Description: A description of your model.
• Machine Learning Service: The service that the model will use (there are others, but for now, we'll select the IBM Watson Machine Learning service: Machine Learning-bg).
• Select model type: In this first attempt, we want to have Model builder help us choose the model type and algorithms, so you can select Automatic.
• Select runtime: Select the runtime environment (just use the Default Spark Scala 2.11 option for now). Runtime environments are the combination of memory, space and CPUs which can effect pricing and cost. You can review the details of various runtimes online (https://dataplatform.cloud.ibm.com/docs/content/wsj/console/environment-runtimes.html).

4. Finally, you can click on Create:

If you miss or skip the step of associating a machine learning service, when you click on Create, you will see the following message:

5. If you do see this message, you will need to click the Associate a Machine Learning service instance link, select a service, and then click on Reload before you can proceed with creating the model.