Using Google Datalab and BigQuery for Image Classification comparison

Part 1: Sample data, simple usage

Google Datalab and BigQuery are useful for image classification projects. We will start with a simple project here. First things first — Google Datalab is used to build Machine Learning (ML) models and runs on Google’s Cloud virtual machine. BigQuery is cloud-based big data analytics web service for processing very large read-only data sets, using SQL-like syntax. Basically, what previously might have been done on a pc or network computer using dedicated resources and an installed database, can now be accessed through a computer with an internet connection. All the heavy lifting and processing is done in the cloud to achieve the same result in a more efficient manner.

Before you begin, ensure that:

• You signed on to a Google Cloud account.
• Google Compute Engine VM is created and active.
• Machine Learning and Dataflow APIs are enabled.
• You have an active project, Datalab and an active notebook.

If you are not sure how to do any of the above, there are several good articles that show you how — here’s one: https://towardsdatascience.com/running-jupyter-notebook-in-google-cloud-platform-in-15-min-61e16da34d52.