IBM Proof of Technology - Data Science Experience

Description:

Use IBM’s Data Science Experience (DSX) and IBM cloud services to create a working cloud-based application from start to finish. Participants will be led through a series of four labs. The first three build upon one another so it is important that they are completed in order.

1. Lab-1 - The first lab will begin with loading raw delimited data into DB2 Warehouse for Cloud and interacting with that data from a Jupyter notebook in DSX with python.

2. Lab-2 - The second lab will leverage Spark machine learning (SparkML) on the loaded data to create categorical predictions using pyspark and a supervised learning model and store the results back to the database.

3. Lab-3 - The third lab will guide participants in creating an R notebook and Shiny UI in DSX using RStudio.

4. Lab-4 - Two labs will be available for Lab 4. Choose one of these labs. One lab will use the Watson Machine Learning capability to create a machine learning model based on the Titanic data set. The model will be deployed in the IBM Cloud, and an application will be built that uses the deployed machine learning model to predict survivability given passenger characteristics. The second lab will use the SPSS Modeler Flow designer to cleanse and prepare the Titanic data set for modeling. A Logistic Regression model will be trained and evaluated to predict survivability given passenger characteristics.

Instructions:

Step 1. Sign up for the IBM Data Science Experience (DSX). We recommend using the Firefox browser.

NOTE: If you already have a DSX account, please skip to Step 2.

Create an account.

1. Go to http://datascience.ibm.com/

2. Click the Sign Up button on the top right

3. Enter your email, click to accept the terms, and click Next. Note: If you already have an IBM id you can click Sign in with your IBM id -- if you don't have a DSX account one will be created for you.

4. On the next page, fill in the corresponding fields and click Create Account

5. You will see the following screen. You need to go to your e-mail account to complete the sign up process.

6. In your e-mail inbox, there should be a message from The Bluemix Team.

7. Open the e-mail and click on Confirm Account.

8. Go back to the DSX session, and enter your DSX user id (e-mail account) and click Continue.

9. Enter your password and click Sign in.

10. Select the IBM Cloud Account (IBM), select the organization (your e-mail), and then select the space (dev) and click Continue.

11. It will take several minutes to set up the account. You will see several screens pop-up (e.g. Preparing Object-Storage, Initializing Environment). Wait until you see the following screen and then click on Get Started.

You are now in the Data Science Experience landing page. Your environment is automatically set up with one Apache Spark instance and 5 GB of object storage. From here you can explore any of the tutorials, videos, sample notebooks, tutorials or articles in the community.

Step 2. Create a project

1. Click on Projects > View All Projects to see a list of your projects. You should only see a default project.

2. Click on the New project icon on the top right of the project list.

3. Type a name for your project. For instance, "DSX Lab". A Spark service and an object storage will be automatically selected as well as a container with a default name. A container is a directory on the object storage. Click on Create.

You are now in your new project where you can create notebooks and data assets as well as add collaborators.

You are now ready to begin Lab-1.