When exploring machine learning algorithms, it can be quite helpful to generate your own data. This gives you complete control and allows for the most exploration of a new technique you might try. It also lets you build trust that your model is working as planned given your assumptions. You've seen this multiple times already in this book up to this point, so it's nothing new. As we develop a linear regression model however, it will be even more instructive since I'm going to work backward through the example.

I will generate data first but show you how I generated the data at the end of the chapter. The goal here is to give you the opportunity to work through building a complex model from a statistical test-first perspective and ultimately show how the generating function was defined and how that affected our work.

The generated data is in the GitHub repo for this book (https://github.com/jcbozonier/Machine-Learning-Test-by-Test) so that you can follow along with the...