Data Science and Engineering at Enterprise Scale

Chapter 5. Classic Machine Learning Examples and Applications

In this chapter we will expand upon our basic understanding of what it means to train a model and begin to apply more sophisticated machine learning concepts. This is only a brief introduction, and you are encouraged to read more deeply into each subject as needed for the problem at hand. We have tried to touch on the most important examples of machine learning, both in the text and in notebook form.

Supervised Learning Models

The most widely used and useful models in machine learning fall under the category of supervised learning. A supervised learning model is simply a model that has been trained on many examples, where a label is associated with a set of features. The most commonly provided label in these models is a binary classification, which is simply an indication that a set of features falls within one of two categories. Models with multiple categories are also possible, but are really just a generalization of the binary classification case, and are not treated in detail here.

The Activation Function: From a Value to a Label

In Chapter 4, we discussed the use of labels and features to train models. The “label” in these cases was actually just a floating-point number. In practice, most incoming data takes on discrete values. The simplest case is a binary label, which indicates whether the data point falls into a particular category or not. We will address how multiple categories are treated a bit later, but ...