Machine learning lab

As part of this lab, you will not be writing any code, you will follow instructions and execute one cell at a time pressing Shift+Enter keys.

To do the lab, you need to open the sagemaker notebook instance and do the step 6 and step 7 from the architecture diagram.

Before you follow instructions in the notebook, let me provide you some background. Imagine that you own a shih-tzu who seems to be popular on the start-up incubator campus you work at. He roams around and everyday you have to take help of campus security to find him. In this lab, you will learn how machine learning can be used to automatically go through hundreds of cameras to locate your dog named Toffee.

Task - Open the notebook instance and follow lab instructions

To open the the notebook instance:

  1. Open AWS Service Catalog.
  2. Choose Provisioned Products from the panel on the left.
  3. Select Provisioned Product with name as App.
  4. Scroll down and look at the Output Key.
  5. You will specify values corresponding to EndpointName and S3Bucket in the notebook.
  6. Click on the link corresponding to Notebook and then click on finding-toffee.ipynb.
  7. Under pre-requisites, specify values for EndpointName and S3Bucket in appropriate cells and then follow the instructions of the lab.

Note that the ML model you are going to use today might return false positive predictions with low probability measure values for random object catagories. However, you will see that the ml model does seem to correctly return high probability measure predictions for the problem we are trying to solve.

After you have completed the task, you can watch the following summary video.