Key concepts of SVR
We will start this section by discussing how support vector machines are used for classification. We will not go into much detail here, leaving a detailed discussion of support vector classification to Chapter 13, Support Vector Machine Classification. But starting with support vector machines for classification will lead nicely to an explanation of SVR.
As I discussed at the beginning of this chapter, support vector machines find the hyperplane that maximizes the margin between classes. When there are only two features present, that hyperplane is just a line. Consider the following example plot:
Figure 8.1 – Support vector machine classification based on two features
The two classes in this diagram, represented by red circles and blue squares, are linearly separable using the two features, x1 and x2. The bold line is the decision boundary. It is the line that is furthest away from border data points for each class, or the maximum...
