Given the set of the following examples as a training set to build a decision tree:
|
EXAMPLE
|
HEIGHT
|
HAIR
|
EYES
|
CLASS
|
|
X1
|
short
|
dark
|
blue
|
+
|
|
X2
|
short
|
blond
|
blue
|
-
|
|
X3
|
tall
|
gray
|
brown
|
-
|
|
X4
|
tall
|
dark
|
blue
|
+
|
|
X5
|
short
|
dark
|
brown
|
-
|
|
X6
|
tall
|
blond
|
blue
|
+
|
Which attribute should be split on first? Why?
Show information gain computations that you used to induce a complete decision tree and draw the tree.
Give the class labels for the following new examples base on your decision tree.
|
Example
|
HEIGHT
|
HAIR
|
EYES
|
CLASS
|
|
X7
|
short
|
gray
|
brown
|
|
|
X8
|
tall
|
dark
|
brown
|
|