Decision Tree Induction - Gene Expression Programming - page 335

9.1 Decision Trees with Nominal Attributes

Describing data using a decision tree is both easy on the eye and an excellent
way of understanding our data. Consider, for instance, the play tennis data
presented in Table 9.1, a famous toy dataset in decision tree induction (Quinlan
1986). This dataset concerns the weather conditions that are suitable for play-
ing tennis. As you can see, there are four different attributes - OUTLOOK,
TEMPERATURE, HUMIDITY, and WINDY - and the decision is whether
to play or not depending on the values of the different attributes. In this
particular case, all four attributes have nominal values. For instance, OUT-
LOOK can be “sunny”, “overcast”, or “rainy”; TEMPERATURE can be “hot”,
“mild”, or “cool”; HUMIDITY can be “high” or “normal”; and WINDY can
be “true” or “false”.

Table 9.1
A small training set with nominal attributes.

OUTLOOK TEMPERATURE HUMIDITY WINDY Play

sunny hot high false No

sunny hot high true No

overcast hot high false Yes

rainy mild high false Yes

rainy cool normal false Yes

rainy cool normal true No

overcast cool normal true Yes

sunny mild high false No

sunny cool normal false Yes

rainy mild normal false Yes

sunny mild normal true Yes

overcast mild high true Yes

overcast hot normal false Yes

rainy mild high true No

A decision tree learned from this dataset is presented in Figure 9.1. And
the classification rules from such a tree are inferred from running all the
paths from the top node to all the leaf nodes. For instance, for the decision
tree of Figure 9.1, there are a total of five different paths or classification
rules. And starting at the most leftward path and continuing towards the right,
they are as follow: