R All-in-One For Dummies

Chapter 2

Decisions, Decisions, Decisions

IN THIS CHAPTER

Finding out about decision trees

Creating a decision tree for the iris dataset

Working with a decision tree for the Acute inflammations dataset from UCI

A decision tree is a graphical way of representing knowledge. As its name implies, it’s a tree-like structure that shows decisions about something, and it’s useful in many fields, from management to medicine.

Think of a decision tree as a way to structure a sequence of questions and possible answers. One prominent use of a decision tree is to show the flow of decision-making to a nontechnical audience.

Decision Tree Components

Figure 2-1 shows a decision tree for classifying irises along with decision tree terminology. If you had a chance to look at Chapter 1 in Book 4, you might recall that the iris dataset (downloaded from the UCI Machine Learning (ML) Repository and designated as iris.uci) consists of 150 rows and 5 columns. The 150 rows represent individual flowers, with 50 each of the setosa, versicolor, and virginica species. The five columns are sepal.length, sepal.width, petal.length, petal.width, and species.

The decision tree is really an upside-down tree, ...