Summary
This was the first purely hands-on chapter in the book. You've connected the theory from the previous chapters with practice. You've built not one, but three fully automated machine learning models. Without any kind of doubt, you should now be able to use TPOT to solve any type of regression problem.
As with most things in data science and machine learning, 90% of the work boils down to data preparation. TPOT can make this percentage even higher because less time is spent designing and tweaking the models. Use this extra time wisely, and get yourself fully acquainted with the dataset. There's no way around it.
In the next chapter, you'll see how to build automated machine learning models for classification datasets. That chapter will also be entirely hands-on. Later, in Chapter 5, Parallel Training with TPOT and Dask, we'll combine both theory and practice.
