2 Introduction to the example problem and Pandas dataframes

In this chapter you’ll learn about the options that you can pick from for a deep learning environment and how to bring tabular structured data into your Python program. You will also learn about the major example that is used throughout this book to demonstrate deep learning for structured data, including details about the dataset used in the major example.

2.1   Development environment options for deep learning

Before you can start a deep learning project you need to have access to an environment that provides the hardware and software stack that you need. If you don’t have direct access to a ready-made deep learning environment in your organization, you still have many choices.

There are several cloud vendors that provide complete deep learning environments for around the cost of a cup of coffee per hour. Each cloud environment has its strengths and weaknesses, with some (Azure and IBM Cloud) emphasizing ease of creating your first project and others (AWS) providing the benefits of scale and incumbency. Appendix A has a more complete description of these options. For now, here is a list of the big cloud vendors who provide deep learning environments:

2.2   Pandas dataframe in Python

2.3   Ingesting CSV files into Pandas dataframes

2.4   The major example: predicting streetcar delays

2.5   Why is a real-world dataset critical for learning about deep learning?

2.6   Format and scope of the input dataset

2.7   Summary