Summary

This chapter introduced key model evaluation measures and techniques so that they will be familiar when we make extensive use of them, and extend them, in the remaining chapters of this book. We examined the very different approaches to evaluation for classification and regression models. We also explored how to use visualizations to improve our analysis of our predictions. Finally, we used pipelines and cross-validation to get reliable estimates of model performance.

I hope this chapter also gave you a chance to get used to the general approach of this book going forward. Although a large number of algorithms will be discussed in the remaining chapters, we will continue to surface the Preprocessing issues we have discussed in the first few chapters. We will discuss the core concepts of each algorithm, of course. But, in a true hands-on fashion, we will also deal with the messiness of real-world data. Each chapter will go from relatively raw data to feature engineering to...