Both the inferential pipelines of our Chaps. 4–6 and the machine-learning pipelines of this chapter and the following one enable us to make predictions: given one or multiple features of an entity (power of a car, temperature and pressure during a rocket launch, photo of a fashion item, ...), we can make an educated guess about another feature of that entity (maximal speed of the car, failure of the rocket, type of the fashion item, ...). But in inferential data analysis, prediction is essentially a byproduct of the models that describe the relationships among the features, whereas in machine-learning, prediction takes center stage. This chapter introduces support-vector machines, possibly the most popular machine-learning approaches to binary classification, and support-vector classifiers, the basic versions of those approaches. Along the way, the chapter compares these methods to the model-based methods of our Chap. 5.