Summary

In this chapter, we covered quite a lot of ground, didn't we!

In short, we learned a lot about different supervised learning algorithms, how to apply them to real datasets, and how to implement everything in OpenCV. We introduced classification algorithms such as k-NN and logistic regression and discussed how they could be used to predict labels as two or more discrete categories. We introduced various variants of linear regression (such as Lasso regression and ridge regression) and discussed how they could be used to predict continuous variables. Last but not least, we got acquainted with the Iris and Boston datasets, two classics in the history of machine learning.

In the following chapters, we will go into much greater depth within these topics, and see some more interesting examples of where these concepts can be useful.

But first, we need to talk about another essential topic of machine learning, feature engineering. Often, data does not come in nicely formatted datasets, and it is our responsibility to represent the data in a meaningful way. Therefore, the next chapter will talk all about representing features and engineering data.