Define the deliverables of the project

In this context, when we talk about deliverables we refer not only to the outputs of the model, but also the deliverables for the project. Think about the following aspects:

• How will the model outputs be used? This can be in many ways: through a dedicated application, via an API, as a module of an existing application, and so on.
• Besides the predictions of the model, what other outputs are needed?
• Are you going to be required to write a report about the results of your model and analysis?
• Will you be required to deliver a presentation?

This is, of course, part of the communication/deployment phase and we will talk about other considerations that we must keep in mind when thinking about the deliverables of the project in the last chapter of the book.

To make what we talked about in this section more concrete, it is time for us to start getting our hands dirty. In the following two sections, we will introduce two of the datasets we will use in the book, along with some hypothetical business problems we are asked to solve. These examples (as well as others in the book) will, of course, be simplified versions of the problems you will encounter in reality, but the essence of doing predictive analytics will be preserved.