Think 2019 Session #3259

Hands On Lab: Introduction to Data Science using IBM Watson Studio.

Description:

Work with IBM's Watson Studio in this workshop to build, train, and test machine learning/deep learning models. Participants will be led through the following three hands-on labs:

1. Lab-1 - The first lab will use Jupyter Notebooks and the XGBoost library to apply machine learning to a classification problem in the healthcare profession. The Watson Machine Learning API will then be used to save and deploy the model.
2. Lab-2 - The second lab will demonstrate Watson Machine Learning capabilites to simplify the building and deployment of machine learning models. The ability to monitor and adjust the deployed model will be demonstrated via the continuous learning capability of Watson Studio.
3. Lab-3 - The third lab will feature the new Watson Studio Neural Network modeler, and Experiment Assistant to build, train, and test a Convolutional Neural Network to classify images.

Before starting the labs, the following prerequisites must be completed.

Prerequisites: Create a Watson Studio project and set up the required services.

Step 1. Log into your Watson Studio account at datascience.ibm.com, then click on the hamburger icon, then Projects, and then View All Projects

Step 2. Click on New Project.

Step 3. Click on Standard

Step 4. Enter the project name (eg. Watson Studio Labs), optionally a description, de-select the Restruct who can be a collaborator checkbox and then click on Add in the Storage section. Note if you have already provisioned cloud object storage (you shouldn't see an Add button) , then just click on the Create button, and skip to Step 9.

Step 5. Click on the Lite plan, and then click on Create.

Step 6. Optionally change the storage name, and then click on Confirm

Step 7. Click on Refresh.

Step 8. The cloud object storage should appear. Now click on Create.

Step 9. Click on the project Settings tab.

Step 10. Scroll down to Associated Services, then select Add service and select Watson.

Step 11. Select the Machine Learning service

Step 12. Select New.

Step 13. Select the Lite plan.

Step 14. Scroll down and click Create, then change the Service name to Machine Learning in the Confirm Creation panel and click Confirm.

Step 15. The Machine Learning service that you created should now appear in Associated Services.

Step 16. Repeat steps 10-15 to create a Spark service.