Preparing IBM Cloud Pak for Data

IBM Spectrum LSF Predictor uses AutoAI in IBM Cloud Pak for Data 4.0 as the training engine. It is required to have Cloud Pak for Data installed and running for LSF Predictor.

AutoAI on private cloud:

1. Install IBM Cloud Pak for Data on premises if not already done.
   For product information, see Cloud Pak for Data.
2. Log in to Cloud Pak for Data as administrator.
3. Create an empty project for IBM Spectrum LSF Predictor.
   1. Click on hamburger drop-down menu.

      

   2. Select Projects.
   3. Click New project.
   4. Select Create an empty project .
   5. Define the project details. Click Create.
4. Record the following information for LSF Predictor settings that you must configure from which a connection can be made to the IBM Cloud Pak for Data instance:
   • IBM Cloud Pak for Data main web console URL
   • IBM Cloud Pak for Data empty project name
   • IBM Cloud Pak for Data administrator user name and password

AutoAI on public cloud:

5. Install IBM Cloud Pak for Data on IBM Cloud® if not already done.
   For product information, see Cloud Pak for Data.
6. Log in to your IBM Cloud account.
7. Provision a Watson Studio service on IBM Cloud.
   Search for "Watson Studio" and select Watson Studio listed under Catalog results.
8. On the Watson Studio service page, select a region, select the Lite (Free) plan, and give your instance a unique service name. Click Create.
9. When the service instance is ready, the Watson Studio page opens. Click Get Started.
10. Under the Work with data heading, click the Create a project link. Click Create an empty project. Give your new project a name.
    Take note of the project name as this name is added on the LSF Predictor settings page.
11. Associate an IBM Cloud Object Storage instance in which to store your data set.
    Under Select storage service, click Add. On the IBM Cloud Object Store service page, leave the service on the Lite tier and click Create. Next, name the service and choose a resource group. Click Create.
12. When your Cloud Object Storage instance is ready, click Refresh on the project page to see your instance under Storage.
13. To create your project, click Create.
14. To add data to your project, click Add to project and select Data.
    You can drag and drop data files under the Load tab, or select browse to navigate to your the location of your data file.
15. To verify that your data file was successfully loaded into your project, select the Assets tab and look for your data file listed under Data assets.
16. Create a deployment space.
    From the main Watson Studio console menu, select View all spaces and click New deployment space +. Fill in the fields, then click Create.
17. In the list of spaces, click on the new deployment space name to open the Space Details page.
18. Copy and save the Space GUID as this space ID is added on the LSF Predictor settings page.
19. Generate an IBM Cloud API key from your IBM Cloud dashboard.
    From the Manage tab, select Access (IAM).

    

20. Select API keys in left-side menu bar.
    Click Create an IBM Cloud API key. Name the API key and click Create.
21. Download the IBM Cloud API key and save it.