There are different categories of machine learning techniques and in this chapter we will see the two most relevant branches—supervised and unsupervised learning, as shown in the following figure:

The supervised and unsupervised learning techniques deal with objects described by features. An example of supervised learning techniques is decision tree learning, and an example of unsupervised technique is k-means. In both cases, the algorithms learn from a set of objects and the difference is their target: supervised techniques predict attributes whose nature is already known and unsupervised techniques identify new patterns.

The supervised learning techniques predict an attribute of the objects. The algorithms learn from a training set of objects whose attribute is known and they predict the attribute of other objects. There are two categories of supervised learning techniques: classification and regression. We talk about classification if the predicted attribute is categoric and about...