Académique Documents
Professionnel Documents
Culture Documents
Recursive Partitioning
Example 2
Simple Tree
Outlook
sunny
overcast
Humidity
high
rainy
Windy
normal
yes
no
Information-Theoretic Approach
To classify an object, a certain information is
needed
I, information
Gain
Gain(A) = I Ires(A)
Entropy
The average amount of information I needed to
classify an object is given by the entropy measure
p(c1)
Residual Information
After applying attribute A, S is partitioned
into subsets according to values v of A
Ires is equal to weighted sum of the
amounts of information for the subsets
Color
green
green
yellow
red
red
red
green
green
yellow
red
green
yellow
yellow
red
Attribute
Outline
dashed
dashed
dashed
dashed
solid
solid
solid
dashed
solid
solid
solid
dashed
solid
dashed
Shape
Dot
no
yes
no
no
no
yes
no
no
yes
no
yes
yes
no
yes
triange
triange
square
square
square
triange
square
triange
square
square
square
square
square
triange
Color
green
green
yellow
red
red
red
green
green
yellow
red
green
yellow
yellow
red
Attribute
Outline
dashed
dashed
dashed
dashed
solid
solid
solid
dashed
solid
solid
solid
dashed
solid
dashed
Shape
Dot
no
yes
no
no
no
yes
no
no
yes
no
yes
yes
no
yes
triange
triange
square
square
square
triange
square
triange
square
square
square
square
square
triange
Data Set:
A set of classified objects
.
Entropy
.
5 triangles
9 squares
class probabilities
.
entropy
Entropy
reduction
by
data set
partitioning
.
.
red
Color?
green
.
yellow
.
.
.
.
.
.
.
.
red
Color?
green
yellow
.
.
Information Gain
.
.
.
.
.
.
red
Color?
green
yellow
.
.
red
Color?
green
.
.
yellow
.
.
red
Color?
green
.
.
yellow
.
.
solid
Outline?
dashed
red
yes
Dot?
Color?
no
green
.
.
yellow
.
.
solid
Outline?
dashed
Decision Tree
.
Color
red
Dot
yes
triangle
yellow
green
square
no
square
Outline
dashed
triangle
solid
square