Summary
This chapter allowed us to explore a very well-known machine learning algorithm: linear regression. We examined the qualities of a feature space that makes it a good candidate for a linear model. We also explored how to improve a linear model, when necessary, with regularization and with transformations. Then, we looked at stochastic gradient descent as an alternative to OLS optimization. We also learned how to add our own classes to a pipeline and how to do hyperparameter tuning.
In the next chapter, we will explore support vector regression.
