Handwritten Digit Recognition Via Classification

Data source: sklearn data sets

Algorithms: support vector machines classifier (SVM)

Tools: IBM Data Science Experience, sklearn

Instructions:

Step I. If you already have an account on IBM's Data Science Experience, go to Step II. If not, follow this tutorial to create an account.

Step II. In the landing page of DSX, create a new project by clicking on the New button on the top right and then click on Project.

Step III. Choose a name for your project (e.g., "Digits"). Choose your default Spark Service and the "Object Storage (Swift API)" option as your Storage Type. Then click Create.

Step IV. Get the Notebook

1. Go to your project and click on the create new notebook icon

2. Click on From URL (3rd tab), choose a name for your notebook (ex: "Digits"), copy and paste this url https://github.com/IBMDataScience/digits/blob/master/Handwritten%20Digits.ipynb into the Notebook URL rectangle and finally click on Create Notebook.

You are now in your new notebook and the rest of the instructions are in there.

NOTE: to execute cells in notebooks select the cell and use Shift+enter