Test-Driven Machine Learning

Chapter 4. Predicting Values with Regression

In this chapter, we'll cover multiple linear regression and how to approach it from a TDD perspective. Unlike the previous chapters, where we developed the actual algorithm using TDD, in this chapter we will explore using a third-party library for the algorithm and TDD building our model. In order to do this, we'll need to find a way to quantify model quality as well as to quantify model assumption violations. We won't have the liberty of checking a data visualization to ensure that our model fits our criteria well.

We will also be using the Python packages statsmodels and pandas, so install those before moving forward in the chapter, using the following commands;

> pip install pandas
> pip install statsmodels ...