09 Decision Tree Induction

Lecture #9
Classification: Decision Tree Induction
Dr. Debasis Samanta

Associate Professor
Department of Computer Science & Engineering
This presentation slides includes
Concept of Decision Tree
Use of Decision Tree to classify data
Basic algorithm to build Decision Tree

Some illustrations
Concept of Entropy
Basic concept of entropy in information theory
Mathematical formulation of entropy
Calculation of entropy of a training set
Decision Tree induction algorithms

ID3
CART
C4.5
CS 40003: Data Analytics 2

Basic Concept
A Decision Tree is an important data structure known to solve many
computational problems
Example 9.1: Binary Decision Tree
A B C f
0 0 0 m0
0 0 1 m1
0 1 0 m2
0 1 1 m3
1 0 0 m4
1 0 1 m5
1 1 0 m6
1 1 1 m7

Basic Concept
In Example 9.1, we have considered a decsion tree where values of any
attribute if binary only. Decision tree is also possible where attributes are
of continuous data type
Example 9.2: Decision Tree with numeric data

Some Characteristics
Decision tree may be n-ary, n 2.
There is a special node called root node.
All nodes drawn with circle (ellipse) are called internal nodes.
All nodes drawn with rectangle boxes are called terminal nodes or leaf
nodes.
Edges of a node represent the outcome for a value of the node.
In a path, a node with same label is never repeated.
Decision tree is not unique, as different ordering of internal nodes can

give different decision tree.

Decision Tree and Classification Task
Decision tree helps us to classify data.
Internal nodes are some attribute
Edges are the values of attributes
External nodes are the outcome of classification
Such a classification is, in fact, made by posing questions starting from

the root node to each terminal node.

Example 9.3 : Vertebrate Classification
Name Body Skin Gives Birth Aquatic Aerial Has Hibernates Class
Temperature Cover Creature Creature Legs
Human Warm hair yes no no yes no Mammal
Python Cold scales no no no no yes Reptile
Salmon Cold scales no yes no no no Fish
Whale Warm hair yes yes no no no Mammal
Frog Cold none no semi no yes yes Amphibian
Komodo Cold scales no no no yes no Reptile
Bat Warm hair yes no yes yes yes Mammal
Pigeon Warm feathers no no yes yes no Bird
Cat Warm fur yes no no yes no Mammal
Leopard Cold scales yes yes no no no Fish
Turtle Cold scales no semi no yes no Reptile
Penguin Warm feathers no semi no yes no Bird
Porcupine Warm quills yes no no yes yes Mammal
Eel Cold scales no yes no no no Fish
Salamander Cold none no semi no yes yes Amphibian
What are the class label of Dragon and Shark?

Example 9.3 : Vertebrate Classification
Suppose, a new species is discovered as follows.
Name Body Skin Gives Aquatic Aerial Has Hibernate Class
Temperature Cover Birth Creature Creature Legs s
Gila Monster cold scale no no no yes yes

?
Decision Tree that can be inducted based on the data (in Example 9.3) is
as follows.

Example 9.3 illustrates how we can solve a classification problem by
asking a series of question about the attributes.
Each time we receive an answer, a follow-up question is asked until we reach a
conclusion about the class-label of the test.
The series of questions and their answers can be organized in the form of
a decision tree
As a hierarchical structure consisting of nodes and edges
Once a decision tree is built, it is applied to any test to classify it.

Definition of Decision Tree
Definition 9.1: Decision Tree
Given a database D = 1 , 2 , . . , , where denotes a tuple, which is

defined by a set of attribute = 1 , 2 , . . , . Also, given a set of classes
C = 1 , 2 , . . , .
A decision tree T is a tree associated with D that has the following properties:
Each internal node is labeled with an attribute Ai
Each edges is labeled with predicate that can be applied to the attribute
associated with the parent node of it
Each leaf node is labeled with class cj

Building Decision Tree
In principle, there are exponentially many decision tree that can be
constructed from a given database (also called training data).
Some of the tree may not be optimum
Some of them may give inaccurate result
Two approaches are known
Greedy strategy
A top-down recursive divide-and-conquer
Modification of greedy strategy

ID3
C4.5
CART, etc.

Built Decision Tree Algorithm
Algorithm BuiltDT
Input: D : Training data set
Output: T : Decision tree
Steps
1. If all tuples in D belongs to the same class Cj
Add a leaf node labeled as Cj
Return // Termination condition
2. Select an attribute Ai (so that it is not selected twice in the same branch)
3. Partition D = { D1, D2, , Dp} based on p different values of Ai in D
4. For each Dk D
Create a node and add an edge between D and Dk with label as the Ais attribute value in Dk
5. For each Dk D
BuildTD(Dk) // Recursive call
6. Stop

Node Splitting in BuildDT Algorithm
BuildDT algorithm must provides a method for expressing an attribute test
condition and corresponding outcome for different attribute type
Case: Binary attribute

This is the simplest case of node splitting
The test condition for a binary attribute generates only two outcomes

Case: Nominal attribute
Since a nominal attribute can have many values, its test condition can be expressed
in two ways:
A multi-way split
A binary split
Muti-way split: Outcome depends on the number of distinct values for the
corresponding attribute
Binary splitting by grouping attribute values

Case: Ordinal attribute
It also can be expressed in two ways:
A multi-way split
A binary split
Muti-way split: It is same as in the case of nominal attribute
Binary splitting attribute values should be grouped maintaining the order property
of the attribute values

Case: Numerical attribute
For numeric attribute (with discrete or continuous values), a test condition can be
expressed as a comparison set
Binary outcome: A >v or A v
In this case, decision tree induction must consider all possible split positions
Range query : vi A < vi+1 for i = 1, 2, , q (if q number of ranges are chosen)
Here, q should be decided a priori
For a numeric attribute, decision tree induction is a combinatorial

optimization problem

Illustration : BuildDT Algorithm
Example 9.4: Illustration of BuildDT Algorithm
Consider a training data set as shown.
Person Gender Height Class

1 F 1.6 S
2 M 2.0 M
3 F 1.9 M
4 F 1.88 M Attributes:
5 F 1.7 S Gender = {Male(M), Female (F)} // Binary attribute
6 M 1.85 M Height = {1.5, , 2.5} // Continuous attribute
7 F 1.6 S
8 M 1.7 S Class = {Short (S), Medium (M), Tall (T)}
9 M 2.2 T
10 M 2.1 T
11 F 1.8 M
12 M 1.95 M Given a person, we are to test in which class s/he belongs
13 F 1.9 M
14 F 1.8 M
15 F 1.75 S

To built a decision tree, we can select an attribute in two different orderings:
<Gender, Height> or <Height, Gender>
Further, for each ordering, we can choose different ways of splitting
Different instances are shown in the following.
Approach 1 : <Gender, Height>


Approach 2 : <Height, Gender>

Example 9.5: Illustration of BuildDT Algorithm
Consider an anonymous database as shown.
A1 A2 A3 A4 Class Is there any clue that enables to

a11 a21 a31 a41 C1 select the best attribute first?
a12 a21 a31 a42 C1
a11 a21 a31 a41 C1 Suppose, following are two
a11 a22 a32 a41 C2 attempts:
a11 a22 a32 a41 C2
A1A2A3A4 [nave]
a12 a22 a31 a41 C1
a11 a22 a32 a41 C2
A3A2A4A1 [Random]
a11 a22 a31 a42 C1 Draw the decision trees in the above-
a11 a21 a32 a42 C2 mentioned two cases.
a11 a22 a32 a41 C2
a12 a22 a31 a41 C1
Are the trees different to classify any test
a12 a22 a31 a42 C1
data?
If any other sample data is added into the
database, is that likely to alter the decision
tree already obtained?

Concept of Entropy

Concept of Entropy
If a point represents a gas molecule,
then which system has the more
entropy?

How to measure? = ?

More ordered Less ordered

less entropy higher entropy
More organized or Less organized or
ordered (less probable) disordered (more probable)

Concept of Entropy
Universe!
What was its entropy value at its starting point?

An Open Challenge!
Roll No. Assignment Project Mid-Sem End-Sem Roll No. Assignment Project Mid-Sem End-Sem
12BT3FP06 89 99 56 91 12BT3FP06 19 59 16 71
10IM30013 95 98 55 93 10IM30013 37 38 25 83
12CE31005 98 96 58 97 12CE31005 38 16 48 97
12EC35015 93 95 54 99 12EC35015 23 95 54 19
12GG2005 90 91 53 98
12GG2005 40 71 43 28
12MI33006 91 93 57 97
12MI33006 61 93 47 97
13AG36001 96 94 58 95
13AG36001 26 64 48 75
13EE10009 92 96 56 96
13EE10009 92 46 56 56
13MA20012 88 98 59 96
13MA20012 88 58 59 66
14CS30017 94 90 60 94
14CS30017 74 20 60 44
14ME10067 90 92 58 95
14ME10067 50 42 38 35
14MT10038 99 89 55 93
14MT10038 29 69 25 33
Two sheets showing the tabulation of marks obtained in a course

are shown.
Which tabulation of marks shows the good performance of the

class?
How you can measure the same?
Entropy and its Meaning
Entropy is an important concept used in Physics in the context of heat
and thereby uncertainty of the states of a matter.
At a later stage, with the growth of Information Technology, entropy
becomes an important concept in Information Theory.
To deal with the classification job, entropy is an important concept,
which is considered as
an information-theoretic measure of the uncertainty contained in a
training data
due to the presence of more than one classes.

Entropy in Information Theory
The entropy concept in information theory first time coined by Claude
Shannon (1850).
The first time it was used to measure the information content in messages.

According to his concept of entropy, presently entropy is widely being used as

a way of representing messages for efficient transmission by
Telecommunication Systems.

Measure of Information Content
People, in general, are information hungry!
Everybody wants to acquire information (from newspaper, library, nature,

fellows, etc.)
Think how a crime detector do it to know about the crime from crime spot and
criminal(s).
Kids annoyed their parents asking questions.
In fact, fundamental thing is that we gather information asking questions (and

decision tree induction is no exception).

We may note that information gathering may be with certainty or uncertainty.

Measure of Information Content
Example 9.6
a) Guessing a birthday of your classmate
1
It is with uncertainty ~
365
1
Whereas guessing the day of his/her birthday is .
7
This uncertainty, we may say varies between 0 to 1, both inclusive.
b) As another example, a question related to event with eventuality (or

impossibility) will be answered with 0 or 1 uncertainty.
Does sun rises in the East? (answer is with 0 uncertainty)
Will mother give birth to male baby? (answer is with uncertainty)
Is there a planet like earth in the galaxy? (answer is with an extreme uncertainty)

Definition of Entropy
Suppose there are m distinct objects, which we want to identify by asking a
series of Yes/No questions. Further, we assume that m is an exact power of 2, say
= 2 , where 1.
Definition 9.2: Entropy
The entropy of a set of m distinct values is the minimum number of yes/no

questions needed to determine an unknown values from these m possibilities.

Entropy Calculation
How can we calculate the minimum number of questions, that is,

entropy?
There are two approaches:
Brute force approach
Clever approach.
Example 9.7: City quiz

Suppose, Thee is a quiz relating to guess a city out of 8 cities, which are as follows:
Bangalore, Bhopal, Bhubaneshwar, Delhi, Hyderabad, Kolkata, Madras, Mumbai
The question is, Which city is called city of joy?

Approach 1: Brute-force search
Brute force approach
We can ask Is it city X?,
if yes stop, else ask next
In this approach, we can ask such questions randomly choosing one city at a time. As a
matter of randomness, let us ask the questions, not necessarily in the order, as they are in
the list.
Q.1: Is the city Bangalore? No
Q.2: Is the city Bhubaneswar? No
Q.3: Is the city Bhopal? No
Q.4: Is the city Delhi? No
Q.5: Is the city Hyderabad? No
Q.6: Is the city Madras? No
Q.7: Is the city Mumbai? No
No need to ask further question! Answer is already out by the Q.7. If asked randomly,
1
each of these possibilities is equally likely with probability 8 . Hence on the average, we
need
(1+2+3+4+5+6+7+7)
= 4.375 questions.
8
Approach 2: Clever approach
Clever approach (binary search)
In this approach, we divide the list into two halves, pose a question for a half
Repeat the same recursively until we get yes answer for the unknown.
Q.1: Is it Bangalore, Bhopal, Bhubaneswar or Delhi? No

Q.2: Is it Madras or Mumbai? No
Q.3: Is it Hyderabad? No
So after fixing 3 questions, we are able to crack the answer.
Note:
Approach 2 is considered to be the best strategy because it will invariably find the answer and
will do so with a minimum number of questions on the average than any other strategy.
Approach 1 occasionally do better (when you are lucky enough!)
It is no coincidence that 8 = 23, and the minimum number of yes/no questions needed is 3.
If m = 16, then 16 = 24 , and we can argue that we need 4 questions to solve the problem.
If m = 32, then 5 questions, m = 256, then 8 questions and so on.

Entropy Calculation
Lemma 9.1: Entropy calculation
The minimum number of yes/no questions needed to identify an unknown
object from = 2 equally likely possible object is n.
If m is not a power of 2, then the entropy of a set of m distinct objects that

are equally likely is log 2

Entropy in Messages
We know that the most conventional way to code information is using binary bits, that
is, using 0s and 1s.
The answer to a question that can only be answered yes/no (with equal probability) can
be considered as containing one unit of information, that is, one bit.
In other words, the unit of information can also be looked at as the amount of
information that can be coded using only 0s and 1s.

Entropy in Messages
Example 9.7: Information coding
If we have two possible objects say male and female, then we use the coding
0 = female
1 = male = 2(= 2 , = 1)
We can encode four possible objects say East, West, North, South using two bits, for example
00 : North
01 : East
10 : West = 4(= 2 , = 2)
11 : South
We can encode eight values say eight different colours, we need to use three bits, such as
000 : Violet
001 : Indigo
010 : Blue
011 : Green
100 : Yellow = 8(= 2 , = 3)
101 : Orange
110 : Red
111 : White
Thus, in general, to code m values, each in a distinct manner, we need n bits such that = 2 .

Entropy in Messages
In this point, we can note that to identify an object, if it is encoded with bits, then we
have to ask questions in an alternative way. For example
Is the first bit 0?
Is the second bit 0?
Is the third bit 0? and so on
Thus, we need n questions, if m objects are there such that = 2 .
The above leads to (an alternative) and equivalent definition of entropy
Definition 9.3: Entropy
The entropy of a set of m distinct values is the number of bits needed to encode all
the values in the most efficient way.

Messages when ( 2 )
In the previous discussion, we have assumed that m, the number of distinct objects is
exactly a power of 2, that is = 2 for some 1 and all m objects are equally likely.
This is mere an assumption to make the discussion simplistic.

In the following we try to redefine the entropy calculation in more general case, that is,
when m 2 and not necessarily m objects are equally probable. Let us consider a
different instance of yes/no question game, which is as follows.
Example 9.8: Name game
There are seven days: Sun, Mon, Tue, Wed, Thu, Fri, Sat.

We are to identify a sequence of 1 such values (each one chosen independently of the others,
that is, repetitions are allowed). Note that if k = 1, it is the type of game, we have already dealt
with.
We denote the minimum number of / questions needed to identify a sequence of unknown

values drawn independently from possibilities as , the entropy in this case.
In other words, is the number of questions required to discriminate amongst distinct
possibilities.

Messages when ( 2 )
Here, = 7 (as stated in the game of sequence of days) and k = 6 (say).
An arbitrary sequence may be {Tue, Thu, Tue, Mon, Sun, Tue}, etc. There are 76 = 117649
possible sequences of six days.
From our previous understanding, we can say that the minimum number of / questions
that is required to identify such a sequence is log 2 11769 = 16.8443.
Since, this is a non integer number, and the number of question should be an integer, we can say
17 questions are required. Thus,
67 = log 2 76
In general,
= log 2
Alternatively, the above can be written as,

log 2 log 2 + 1
Or
1
log 2 log 2 +


Entropy of Messages when ( 2 )

Note that here is the average number of questions needed to determine each of the
values in a sequence of k values. By choosing a large enough value of k, that is, a long
1
enough sequence, the value of can be made as small as we wish. Thus, the average
number of questions required to determine each value can be made arbitrarily close to
log 2 . This is evident from our earlier workout, for example, tabulated below, for m = 7.
= log 2
k log 2 No. Q .

6 117649 16.84413 17 2.8333
21 58.95445 59 2.8095
1000 2807.3549 2808 2.8080
.. .. .. .. ..
No. Q = Number of questions
. . 7 7
Note that log 2 7 2.8074 and log 2 7. Further, = i.e. = log 2 7 (is

independent of k and is a constant!)
Entropy of Messages when ( 2 )
Lemma 9.4: Entropy Calculation
The entropy of a set of m distinct objects is log 2 even when m is not exactly a
power of 2.
We have arrived at a conclusion that E = log 2 for any value of m, irrespective of

weather it is a power of 2 or not.
Note: E is not necessarily be an integer always.
Next, we are to have our observation, if all m objects are not equally probable.
Suppose, denotes the frequency with which the of the m objects occurs,
where 0 1 for all such that

= 1
=1

Discriminating amongst m values ( 2 )
Example 9.8: Discriminating among objects
Suppose four objects , , which occur with frequencies 12, 14, 1
8
1
and 8,
respectively.
1 1 1
Thus, in this example, = 4 and 1 = , 2 = , 3 = and 4 = .
2 4 8
1
8
Using standard 2-bit encoding, we can represent them as = 00, = 01, = 10,
= 11.
Also, we can follow variable length coding (also called Huffman coding) as an
improved way of representing them.
1 1 1 1
The Huffman coding of , , with their frequencies 2, 4, 8 and 8 are shown
below.
A=1
B = 01
C = 001
D = 000

With the above representation say, if A is to be identified, then we need to
examine only one question, for B it is 2 and for C and D both, it is 3.
Thus, on the average, we need
1 1 1 1
1 + 4 2 + 8 3 + 8 3 = 1.75
2
This is the number of yes/no questions to identify any one of the four objects,
whose frequency of occurrences are not uniform.
This is simply in contrast to 2-bit encoding, where we need 2-bits (questions)

on the average.

It may be interesting to note that even with variable length encoding, there are
several ways of encoding. Few of them are given below.
1) = 0 2) = 01 3) = 101
= 11 =1 = 001
= 100 = 001 = 10011
= 101 = 000 = 100001
The calculation of entropy in the observed cases can be obtained as:

1) 1.75 2) 2 3) 3.875
Anyway, key to finding the most efficient way of encoding is to assign a
smallest number of bits to the object with highest frequency and so on.
The above observation is also significant in the sense that it provides a
systematic way of finding a sequence of well-chosen question in order to
identify an object at a faster rate.

Information Content
Based on the previous discussion we can easily prove the following lemma.
Lemma 9.3: Information content

If an object occurs with frequency p, then the most efficient way to represent it
with log 2 (1) bits.
Example 9.9: Information content

1
A which occurs with frequency is represented by 1-bit, B which occurs with
2
1 1
frequency 4 represented by 2-bits and both C and D which occurs with frequency 8
are represented by 3 bits each.

Entropy Calculation
We can generalize the above understanding as follows.
If there are m objects with frequencies 1 , 2 ., , then the average number of
bits (i.e. questions) that need to be examined a value, that is, entropy is the frequency
of occurrence of the value multiplied by the number of bits that need to be
determined, summed up values of from 1 to m.
Theorem 9.4: Entropy calculation
If pi denotes the frequencies of occurrences of m distinct objects, then the entropy

E is

= log(1 ) = 1
=1 =1
Note:
1
If all are equally likely, then = and = log 2 ; it is the special case.

Entropy of a Training Set
If there are k classes 1 , 2 ., and for = 1 denotes the number of
occurrences of classes divided by the total number of instances (i.e., the frequency
of occurrence of ) in the training set, then entropy of the training set is denoted by

= log 2
=1
Here, E is measured in bits of information.
Note:
The above formula should be summed over the non-empty classes only, that is, classes
for which 0
E is always a positive quantity
E takes its minimum value (zero) if and only if all the instances have the same class
(i.e., the training set with only one non-empty class, for which the probability 1).
Entropy takes its maximum value when the instances are equally distributed among k
possible classes. In this case, the maximum value of E is 2 .

Entropy of a Training Set
Example 9.10: OPTH dataset
Consider the OTPH data shown in the following table with total 24 instances in it.
Age Eye sight Astigmatic Use Type Class

1 1 1 1 3
1 1 1 2 2
1 1 2 1 3
1 1 2 2 1
1 2 1 1 3
1 2 1 2 2
1 2 2 1 3
1 2 2 2 1
2 1 1 1 3
2 1 1 2 2
2 1 2 1 3 A coded
2 1 2 2 1
forms for all
2 2 1 1 3 values of
2 2 1 2 2
2 2 2 1 3 attributes
2 2 2 2 3 are used to
3 1 1 1 3
3 1 1 2 3 avoid the
3 1 2 1 3 cluttering in
3 1 2 2 1 the table.
3 2 1 1 3
3 2 1 2 2
3 2 2 1 3
CS 40003: Data Analytics 3 2 2 2 3 48
Entropy of a training set
Specification of the attributes are as follows.
Age Eye Sight Astigmatic Use Type

1: Young 1: Myopia 1: No 1: Frequent
2: Middle-aged 2: Hypermetropia 2: Yes 2: Less
3: Old
Class: 1: Contact Lens 2:Normal glass 3: Nothing
In the OPTH database, there are 3 classes and 4 instances with class 1, 5 instances
with class 2 and 15 instances with class 3. Hence, entropy E of the database is:
4 4 5 5 15 15
= log 2 log 2 log 2 = 1.3261
24 24 24 24 24 24

Note:
The entropy of a training set implies the number of yes/no questions, on the average,
needed to determine an unknown test to be classified.
It is very crucial to decide the series of questions about the value of a set of attribute,
which collectively determine the classification. Sometimes it may take one question,
sometimes many more.
Decision tree induction helps us to ask such a series of questions. In other words, we
can utilize entropy concept to build a better decision tree.
How entropy can be used to build a decision tree is our next topic of discussion.

Decision Tree Induction Techniques
Decision tree induction is a top-down, recursive and divide-and-conquer
approach.
The procedure is to choose an attribute and split it into from a larger training
set into smaller training sets.
Different algorithms have been proposed to take a good control over
1. Choosing the best attribute to be splitted, and
2. Splitting criteria
Several algorithms have been proposed for the above tasks. In this lecture, we
shall limit our discussions into three important of them
ID3
C 4.5
CART

Algorithm ID3

ID3: Decision Tree Induction Algorithms
Quinlan [1986] introduced the ID3, a popular short form of Iterative
Dichotomizer 3 for decision trees from a set of training data.
In ID3, each node corresponds to a splitting attribute and each arc is a

possible value of that attribute.
At each node, the splitting attribute is selected to be the most informative

among the attributes not yet considered in the path starting from the root.

Algorithm ID3
In ID3, entropy is used to measure how informative a node is.
It is observed that splitting on any attribute has the property that average entropy
of the resulting training subsets will be less than or equal to that of the previous
training set.
ID3 algorithm defines a measurement of a splitting called Information Gain

to determine the goodness of a split.
The attribute with the largest value of information gain is chosen as the splitting
attribute and
it partitions into a number of smaller training sets based on the distinct values of
attribute under split.

Defining Information Gain
We consider the following symbols and terminologies to define
information gain, which is denoted as .
D denotes the training set at any instant
|D| denotes the size of the training set D
E(D) denotes the entropy of the training set D
The entropy of the training set D
E(D) = -=1 2 (
where the training set D has 1 , 2 , , , the k number of distinct classes and
, 0< 1 is the probability that an arbitrary tuple in D belongs to class (i =

1, 2, , k).

pi can be calculated as
|, |
=
||
where , is the set of tuples of class in D.
Suppose, we want to partition D on some attribute A having m distinct values

{1 , 2 , , }.
Attribute can be considered to split D into m partitions {1 , 2 , , },

where (j = 1,2, ,m) contains those tuples in D that have outcome of A.

Definition 9.4: Weighted Entropy
The weighted entropy denoted as () for all partitions of D with respect to A is

given by:
| |
() =
=1 E( )
||
| |
Here, the term denotes the weight of the j-th training set.
||
More meaningfully, () is the expected information required to classify a tuple

from D based on the splitting of A.

Our objective is to take A on splitting to produce an exact classification (also
called pure), that is, all tuples belong to one class.
However, it is quite likely that the partitions is impure, that is, they contain
tuples from two or more classes.
In that sense, () is a measure of impurities (or purity). A lesser value of
() implying more power the partitions are.
Definition 9.5: Information Gain
Information gain, , of the training set D splitting on the attribute A is given

by
, = E(D) - ()
In other words, , gives us an estimation how much would be gained by

splitting on A. The attribute A with the highest value of should be chosen as the
splitting attribute for D.

Information Gain Calculation
Example 9.11 : Information gain on splitting OPTH
Let us refer to the OPTH database discussed in Slide #48.

Splitting on Age at the root level, it would give three subsets 1 , 2 and 3 as
shown in the tables in the following three slides.
The entropy (1 ), (2 ) and (3 ) of training sets 1 , 2 and 3 and
corresponding weighted entropy (1 ), (2 ) and (3 ) are also
shown alongside.
The Information gain (, ) is then can be calculated as 0.0394.
Recall that entropy of OPTH data set, we have calculated as E(OPTH) =

1.3261
(see Slide #49)
59
CS 40003: Data Analytics
Information Gain Calculation
Example 9.11 : Information gain on splitting OPTH
Training set: 1 (Age = 1)
Age Eye-sight Astigmatism Use type Class

1 1 1 1 3
1 1 1 2 2 2 2 2 2
(1 ) = 8 2 (8) 8 2 (8)
1 1 2 1 3
4 4
1 1 2 2 1 8 2 (8) = 1.5
1 2 1 1 3 8
(1 ) = 1.5 = 0.5000
1 2 1 2 2 24
1 2 2 1 3
1 2 2 2 1
60
CS 40003: Data Analytics
Calculating Information Gain

2 1 1 1 3
2 1 1 2 2
2 1 2 1 3
2 1 2 2 1
2 2 1 1 3
2 2 1 2 2
2 2 2 1 3
2 2 2 2 3
1 1 2 2 5 5
(2 ) = 2 ( ) 2( ) 2 ( )
8 8 8 8 8 8
= 1.2988
8
(2 ) = 1.2988 = 0.4329
CS 40003: Data Analytics 24 61
Calculating Information Gain

3 1 1 1 3
1 1 1 1
3 1 1 2 3 (3 ) = 8 2 (8) 8 2(8)
6 6
3 1 2 1 3 8 2 (8) = 1.0613
3 1 2 2 1 8
(3 ) = 1.0613 = 0.3504
3 2 1 1 3 24
3 2 1 2 2
3 2 2 1 3
3 2 2 2 3
(, ) = 1.3261 (0.5000 + 0.4329 + 0.3504) = 0.0394

Information Gains for Different Attributes
In the same way, we can calculate the information gains, when splitting the
OPTH database on Eye-sight, Astigmatic and Use Type. The results are
summarized below.
Splitting attribute: Age
, = 0.0394
Splitting attribute: Eye-sight

, = 0.0395
Splitting attribute: Astigmatic

Astigmatic , = 0.3770
Splitting attribute: Use Type

Use Type , = 0.5488

Decision Tree Induction : ID3 Way
The ID3 strategy of attribute selection is to choose to split on the attribute that
gives the greatest reduction in the weighted average entropy
The one that maximizes the value of information gain
In the example with OPTH database, the larger values of information gain is
Use Type , = 0.5488
Hence, the attribute should be chosen for splitting is Use Type.
The process of splitting on nodes is repeated for each branch of the evolving
decision tree, and the final tree, which would look like is shown in the following
slide and calculation is left for practice.

Decision Tree Induction : ID3 Way
= .

Age Eye-sight Use Type Astigmatic
= 0.0394 = 0.395 = 0.0394 = 0.3770
1 (1) 2 (2)
Age Eye Ast Use Class Age Eye Ast Use Class
1 1 1 2 2
1 1 1 1 3
1 1 2 2 1
1 1 2 1 3
1 2 1 2 2
1 2 1 1 3
1 2 2 2 1
(1 ) =?
1 2 2 1 3 (2 ) =?
2 1 1 1 3 2 1 1 2 2
2 2 1 1 3 2 1 2 2 1
2 2 2 1 3 2 2 1 2 2
3 1 1 1 3 3 1 1 2 3
3 1 2 1 3 3 1 2 2 3
3 2 1 1 3 3 2 1 2 2
3 2 2 1 3 3 2 2 2 3
? ? ? ? ? ?
Age Eye-
Astigmatic Age Eye-
sight Astigmatic
=? sight
CS 40003: Data Analytics =? =? =? =? =? 65
Frequency Table : Calculating
Calculation of entropy for each table and hence information gain for a
particular split appears tedious (at least manually)!
As an alternative, we discuss a short-cut method of doing the same using a
special data structure called Frequency Table.
Frequency Table: Suppose, = {1 , 2 , , } denotes an attribute with

attribute values in it. For a given database , there are a set
of say = {1 , 2 , }. Given this, a frequency table will look
like as follows.

Frequency Table : Calculating

1 2
1
Number of rows = Number
2 of classes
Number of columns =
Class
Number of attribute values

= Frequency of for
class
Assume that = , the number of total instances of .

Calculation of using Frequency Table
Example 9.12 : OTPH Dataset
With reference to OPTH dataset, and for the attribute Age, the frequency table would
look like
Age=1 Age=2 Age=3 Row Sum
Class 1 2 1 1 4
Class 2 2 2 1 5
N=24
Class 3 4 5 6 15
Column
8 8 8 24
Sum
Column Sums

The weighted average entropy () then can be calculated from the
frequency table following the
Calculate = log 2 for all = 1,2, ,
(Entry Sum) = 1, 2, , and 0
Calculate = log 2 for all = 1, 2, ,
(Column Sum) in the row of column sum
Calculate = ( + )/
Example 9.13: OTPH Dataset

For the frequency table in Example 9.12, we have

= 2 log 2 + 1 log 1 + 1 log 1 + 2 log 2 + 2 log 2 + 1 log 1 + 4 log 4 + 5 log 5
+ 6 log 6
= 8 log 8 + 8 log 8 + 8 log 8
= .
Proof of Equivalence
In the following, we prove the equivalence of the short-cut of entropy
calculation using Frequency Table.
Splitting on an attribute with values produces subsets of the training
dataset (of size = ). The subset ( = 1,2, , ) contains all
the instances for which the attribute takes its value. Let denotes
the number of instances in the subset. Then

=
=1
Let denotes the number of instances for which the classification is

and attribute takes its value. Then
=
=1
Denoting as the entropy of the subset, we have

= log 2

=1
Therefore, the weighted average entropy of the splitting attribute is given by

= .

=1

= . . log 2

=1 =1

= . log 2

=1 =1

= . log 2 + . log 2

=1 =1 =1 =1
1

= . log 2 + . log 2

=1 =1 =1

=
=1

= . log 2 + log 2
=1 =1 =1
= ( + )
where = . log 2 (Entries sum)

=1 =1
= log 2 ( )
=1
Hence, the equivalence is proved.
Limiting Values of Information Gain
The Information gain metric used in ID3 always should be positive or
zero.
It is always positive value because information is always gained (i.e.,

purity is improved) by splitting on an attribute.
On the other hand, when a training set is such that if there are classes,
and the entropy of training set takes the largest value i.e., log 2 (this
occurs when the classes are balanced), then the information gain will be
zero.

Limiting Values of Information Gain
Example 9.14: Limiting values of Information gain
Consider a training set shown below.

Data set Table A X Table X
X Y Class 1 2 3 4
1 1 A A 1 1 1 1
1 2 B Frequency table of X
B 1 1 1 1
2 1 A . 2 2 2 2
2 2 B
Y Y Table Y
3 2 A
1 2
3 1 B
A 2 2
4 2 A Frequency table of Y
B 2 2
4 1 B
. 4 4

Limiting values of Information Gain
Entropy of Table A is
1 1 1 1
= log log = log 2 = 1 (The maximum entropy).
2 2 2 2
In this example, whichever attribute is chosen for splitting, each of the

branches will also be balanced thus each with maximum entropy.
In other words, information gain in both cases (i.e., splitting on X as well as
Y) will be zero.
Note:
The absence of information gain does not imply that there is no profit for
splitting on the attribute.
Even if it is chosen for splitting, ultimately it will lead to a final decision tree
with the branches terminated by a leaf node and thus having an entropy of
zero.
Information gain can never be a negative value.
Splitting of Continuous Attribute Values
In the foregoing discussion, we assumed that an attribute to be splitted is
with a finite number of discrete values. Now, there is a great deal if the
attribute is not so, rather it is a continuous-valued attribute.
There are two approaches mainly to deal with such a case.
1. Data Discretization: All values of the attribute can be discretized into a

finite number of group values and then split point can be decided at each
boundary point of the groups.
1 2 3
So, if there are of discrete values, then we have ( + 1) split

points.

Splitting of Continuous attribute values
2. Mid-point splitting: Another approach to avoid the data discretization.
It sorts the values of the attribute and take the distinct values only in it.
Then, the mid-point between each pair of adjacent values is considered as a
split-point.
1 2 3 +1
Here, if n-distinct values are there for the attribute , then we choose 1
split points as shown above.
+ (+1)
For example, there is a split point = in between and (+1)
2
For each split-point, we have two partitions: > , and finally the
point with maximum information gain is the desired split point for that
attribute.
Algorithm CART

CART Algorithm
It is observed that information gain measure used in ID3 is biased towards test
with many outcomes, that is, it prefers to select attributes having a large number
of values.
L. Breiman, J. Friedman, R. Olshen and C. Stone in 1984 proposed an algorithm
to build a binary decision tree also called CART decision tree.
CART stands for Classification and Regression Tree
In fact, invented independently at the same time as ID3 (1984).
ID3 and CART are two cornerstone algorithms spawned a flurry of work on decision
tree induction.
CART is a technique that generates a binary decision tree; That is, unlike ID3,
in CART, for each node only two children is created.
ID3 uses Information gain as a measure to select the best attribute to be splitted,
whereas CART do the same but using another measurement called Gini index . It
is also known as Gini Index of Diversity and is denote as .

Gini Index of Diversity
Definition 9.6: Gini Index
Suppose, D is a training set with size |D| and = 1 , 2 , , be the set of k

classifications and = 1 , 2 , , be any attribute with m different values of
it. Like entropy measure in ID3, CART proposes Gini Index (denoted by G) as the
measure of impurity of D. It can be defined as follows.

= 1 2
=1
where is the probability that a tuple in D belongs to class and can be
estimated as
| , |
=

where | , | denotes the number of tuples in D with class .

Note
measures the impurity of data set D.
The smallest value of is zero
which it takes when all the classifications are same.
1
It takes its largest value = 1

when the classes are evenly distributed between the tuples, that is the frequency of
1
each class is .

Definition 9.7: Gini Index of Diversity
Suppose, a binary partition on A splits D into 1 and 2 , then the weighted

average Gini Index of splitting denoted by () is given by
1 2
= . 1 + . (2 )

This binary partition of D reduces the impurity and the reduction in impurity is
measured by
, =

Gini Index of Diversity and CART
This , is called the Gini Index of diversity.
It is also called as impurity reduction.
The attribute that maximizes the reduction in impurity (or equivalently, has
the minimum value of ()) is selected for the attribute to be splitted.

n-ary Attribute Values to Binary Splitting
The CART algorithm considers a binary split for each attribute.
We shall discuss how the same is possible for attribute with more than two
values.
Case 1: Discrete valued attributes

Let us consider the case where A is a discrete-valued attribute having m
discrete values 1 , 2 , , .
To determine the best binary split on A, we examine all of the possible subsets
say 2 of A that can be formed using the values of A.
Each subset 2 can be considered as a binary test for attribute A of the

form " ? .

Thus, given a data set D, we have to perform a test for an attribute value A
like D
Yes No
1 2
This test is satisfied if the value of A for the tuples is among the values listed
in .
If A has m distinct values in D, then there are 2 possible subsets, out of
which the empty subset { } and the power set {1 , 2 , , } should be
excluded (as they really do not represent a split).
Thus, there are 2 2 possible ways to form two partitions of the dataset D,
based on the binary split of A.
Case2: Continuous valued attributes
For a continuous-valued attribute, each possible split point must be taken into
account.
The strategy is similar to that followed in ID3 to calculate information gain

for the continuous valued attributes.
According to that strategy, the mid-point between ai and ai+1 , let it be vi, then
1 2 +1
Yes No
+ +1
= 1 2
2

Each pair of (sorted) adjacent values is taken as a possible split-point say .
1 is the set of tuples in D satisfying and 2 in the set of tuples in D

satisfying > .
The point giving the minimum Gini Index is taken as the split-point of
the attribute A.
Note
The attribute A and either its splitting subset (for discrete-valued splitting
attribute) or split-point (for continuous valued splitting attribute) together
form the splitting criteria.

CART Algorithm : Illustration
Example 9.15 : CART Algorithm
Suppose we want to build decision tree for the data set EMP as given in the table
below.
Age Tuple# Age Salary Job Performance Select
Y : young 1 Y H P A N
M : middle-aged
2 Y H P E N
O : old
3 M H P A Y
Salary
4 O M P A Y
L : low
M : medium 5 O L G A Y
H : high 6 O L G E N
Job 7 M L G E Y
G : government 8 Y M P A N
P : private
9 Y L G A Y
Performance 10 O M G A Y
A : Average
11 Y M G E Y
E : Excellent
12 M M P E Y
Class : Select
13 M H G A Y
Y : yes
N : no 14 O M P E N

For the EMP data set,
2
= 1 2
=1
2 2
9 5
=1 +
14 14
= .
Now let us consider the calculation of for Age, Salary, Job and
Performance.

Attribute of splitting: Age
The attribute age has three values, namely Y, M and O. So there are 6 subsets, that should
be considered for splitting as:
{} {} {} {, } {, } {, }
, , , , ,
1 2 3 4 5 6
2 2 2 2
5 3 2 9 6 8
1 = 1 + 1 = .
14 5 5 14 14 14
2 = ?
3 = ? Age {O}
4 = 3 Yes No
5 = 2
{O} {Y,M}
6 = 1
The best value of Gini Index while splitting attribute Age is 3, = .

Attribute of Splitting: Salary
The attribute salary has three values namely L, M and H. So, there are 6 subsets,
that should be considered for splitting as:
{} {, } {} {, } {} {, }
, , , , ,
1 2 3 4 5 6
Salary {H}
1 = 2 = 0.3000
Yes No
3 = 4 = 0.3150
5 = 6 = 0.4508 {H} {L,M}
(5,6), = 0.4592 0.4508 = 0.0084

Attribute of Splitting: job
Job being the binary attribute , we have

7 7
= 1 + (2 )
14 14
7 3 2 4 2 7 6 2 1 2
= 1 + 1 =?
14 7 7 14 7 7
, = ?

Attribute of Splitting: Performance
Job being the binary attribute , we have

= ?
, = ?
Out of these , gives the maximum value and hence, the attribute
Salary would be chosen for splitting subset {, } or {}.
Note:
It can be noted that the procedure following information gain calculation (i.e.
(, )) and that of impurity reduction calculation ( i.e. , ) are near
about.

Calculating using Frequency Table
We have learnt that splitting on an attribute gives a reduction in the average
Gini Index of the resulting subsets (as it does for entropy).
Thus, in the same way the average weighted Gini Index can be calculated
using the same frequency table used to calculate information gain , ,
which is as follows.
The G( ) for the subset

2

= 1
| |
=1

Calculating using Frequency Table
The average weighted Gini Index, ( ) (assuming that attribute has m distinct
values is)

| |
= .
|1 |
=1
2
| | | |
= .
|| || | |
=1 =1 =1

1 1
=1 . 2
||
=1 =1
The above gives a formula for m-attribute values; however, it an be fine tuned to
subset of attributes also.

Illustration: Calculating using Frequency Table
Example 9.16 : Calculating using frequency table of OPTH
Let us consider the frequency table for OPTH database considered earlier. Also
consider the attribute 1 with three values 1, 2 and 3. The frequency table is
shown below.
1 2 3
Class 1 2 1 1
Class 2 2 2 1
Class 3 4 5 6
Column sum 8 8 8

Now we can calculate the value of Gini Index with the following steps:
1. For each non-empty column, form the sum of the squares of the values in the
body of the table and divide by the column sum.
2. Add the values obtained for all columns and divided by ||, the size of the
database.
3. Subtract the total from 1.
As an example, with reference to the frequency table as mentioned just.
22 +22 +42 12 +22 +52

1 = 1 = = 3.0 1 = 2 = = 3.75
24 24
12 +12 +62 1+3.75+4.75
1 = 3 = = 4.75 So, 1 = 1 = 0.5208
24 24

Thus, the reduction in the value of Gini Index on splitting attribute 1 is
1 , = 0.5382 0.5208 = 0.0174
where = 0.5382
The calculation can be extended to other attributes in the OTPH database and is
left as an exercise.
?
Decision Trees with ID3 and CART Algorithms
Example 9.17 : Comparing Decision Trees of EMP Data set
Compare two decision trees obtained using ID3 and CART for the EMP dataset.
The decision tree according to ID3 is given for your ready reference (subject to
the verification)
Y Age
O
Job Y Performance
P G A E
N Y Y N
Decision Tree using ID3
?
Decision Tree using CART

Algorithm C4.5

Algorithm C 4.5 : Introduction
J. Ross Quinlan, a researcher in machine learning developed a decision tree
induction algorithm in 1984 known as ID3 (Iterative Dichotometer 3).
Quinlan later presented C4.5, a successor of ID3, addressing some limitations

in ID3.
ID3 uses information gain measure, which is, in fact biased towards splitting
attribute having a large number of outcomes.
For example, if an attribute has distinct values for all tuples, then it would
result in a large number of partitions, each one containing just one tuple.
In such a case, note that each partition is pure, and hence the purity measure of the
partition, that is = 0

Algorithm C4.5 : Introduction
Example 9.18 : Limitation of ID3
In the following, each tuple belongs to a unique class. The splitting on A is shown.

1
= . = .0 = 0

=1 =1
Thus, , = is maximum in such a situation.

Algorithm: C 4.5 : Introduction
Although, the previous situation is an extreme case, intuitively, we can infer
that ID3 favours splitting attributes having a large number of values
compared to other attributes, which have a less variations in their values.
Such a partition appears to be useless for classification.
This type of problem is called overfitting problem.
Note:
Decision Tree Induction Algorithm ID3 may suffer from overfitting problem.

Algorithm: C 4.5 : Introduction
The overfitting problem in ID3 is due to the measurement of information gain.
In order to reduce the effect of the use of the bias due to the use of
information gain, C4.5 uses a different measure called Gain Ratio, denoted as
.
Gain Ratio is a kind of normalization to information gain using a split

information.
Algorithm: C 4.5 : Gain Ratio
Definition 9.8: Gain Ratio
The gain ratio can be defined as follows. We first define split information
as

= . log

=1
Here, m is the number of distinct values in A.

(,)
The gain ratio is then defined as , = , where (, ) denotes the

information gain on splitting the attribute A in the dataset D.

Physical Interpretation of (D)
(D)
The value of split information depends on

the number of (distinct) values an attribute has and
how uniformly those values are distributed.
In other words, it represents the potential information generated by splitting a

data set D into m partitions, corresponding to the m outcomes of on attribute A.
Note that for each outcome, it considers the number of tuples having that
outcome with respect to the total number of tuples in D.

Physical Interpretation of (D )
Example 9.18 : (D)
To illustrate (D), let us examine the case where there are 32 instances and
splitting an attribute A which has 1 , 2 , 3 and 4 sets of distinct values.
Distribution 1 : Highly non-uniform distribution of attribute values
1 2 3 4
Frequency 32 0 0 0
32 32
(D) = 2 ( ) = 2 1 = 0
32 32
Distribution 2
1 2 3 4
Frequency 16 16 0 0
16 16 16 16
(D) = (
2 32) 2 32)
( = 2 2 = 1
32 32

Distribution 3
1 2 3 4
Frequency 16 8 8 0
16 16 8 8 8 8
(D) = (
2 32) (
2 32 ) 2 32 )
( = 1.5
32 32 32
Distribution 4
1 2 3 4
Frequency 16 8 4 4
(D) = 1.75
Distribution 5: Uniform distribution of attribute values
1 2 3 4
Frequency 8 8 8 8
8 8 1
(D) = ( 32 2 (32))*4 = 2 (4)= 2.0

In general, if there are m attribute values, each occurring equally frequently,
then the split information is 2 .
Based on the Example 9.18, we can summarize our observation on split
information as under:
Split information is 0 when there is a single attribute value. It is a trivial case and
implies the minimum possible value of split information.
For a given data set, when instances are uniformly distributed with respect to the
attribute values, split information increases as the number of different attribute
values increases.
The maximum value of split information occur when there are many possible
attribute values, all are equally frequent.
Note:
Split information varies between 0 and log2m (both inclusive)

Physical Interpretation of ,
Information gain signifies how much information will be gained on
partitioning the values of attribute A
Higher information gain means splitting of A is more desirable.

On the other hand, split information forms the denominator in the gain ratio
formula.
This implies that higher the value of split information is, lower the gain ratio.
In turns, it decreases the information gain.
Further, information gain is large when there are many distinct attribute
values.
When many distinct values, split information is also a large value.
This way split information reduces the value of gain ratio, thus resulting a
balanced value for information gain.
Like information gain (in ID3), the attribute with the maximum gain ratio is
selected as the splitting attribute in C4.5.

The frequency table can be used to calculate the gain ratio for a given data set and
an attribute.

We have already learned the calculation of information gain using Frequency

Table.
To calculate gain ratio, in addition to information gain, we are also to calculate
split information.
This split information can be calculated from frequency table as follows.
For each non-zero column sum say contribute |Dj | for the j-th column (i.e., the
j-th value of the attribute). Thus the split information is

(D) = m 2 ||
=1 |D|
If there are m-columns in the frequency table.
Practice:
Using Gain ratio as the measurement of splitting attributes, draw the decision
trees for OPTH and EMP data sets. Give calculation of gain ratio at each node.
Summary of Decision Tree Induction Algorithms
We have learned the building of a decision tree given a training data.
The decision tree is then used to classify a test data.
For a given training data D, the important task is to build the decision tree so
that:
All test data can be classified accurately
The tree is balanced and with as minimum depth as possible, thus the classification
can be done at a faster rate.
In order to build a decision tree, several algorithms have been proposed.

These algorithms differ from the chosen splitting criteria, so that they satisfy
the above mentioned objectives as well as the decision tree can be induced
with minimum time complexity. We have studied three decision tree induction
algorithms namely ID3, CART and C4.5. A summary of these three
algorithms is presented in the following table.

Table 11.6
Algorithm Splitting Criteria Remark
ID3 Information Gain The algorithm calculates
, = (D) ( ,D) for all in D
Where and choose that attribute
= Entropy of D (a which has maximum
measure of uncertainty) = ( ,D).
=1 log 2
where D is with set of k classes
1 , 2 , , and = , ; The algorithm can handle
| |
|| both categorical and
Here, , is the set of tuples numerical attributes.
with class in D.
(D) = Weighted average
entropy when D is partitioned It favors splitting those
on the values of attribute A = attributes, which has a
| |

=1 ( ) large number of distinct
||
Here, m denotes the distinct values.
values of attribute A.

CART Gini Index The algorithm calculates
, = (D) all binary partitions for
where all possible values of
= Gini index (a measure of attribute A and choose
impurity) that binary partition
= 1 =1 2 which has the maximum
|, | , .
Here, = || and D is with k
number of classes and The algorithm is
| | | | computationally very
GA(D) = ||1 (1 ) + ||2 (2 ),
expensive when the
when D is partitioned into two attribute A has a large
data sets 1 and 2 based on number of values.
some values of attribute A.

C4.5 Gain Ratio The attribute A with
, maximum value of
, =
(D) , is selected for
where splitting.
, = Information gain of
D (same as in ID3, and Splitting information is a
(D) = splitting information kind of normalization, so
| | | | that it can check the
=
=1 2
|| || biasness of information
when D is partitioned into 1 , gain towards the
2 , , partitions choosing attributes with a
corresponding to m distinct large number of distinct
attribute values of A. values.
In addition to this, we also highlight few important characteristics

of decision tree induction algorithms in the following.

Notes on Decision Tree Induction algorithms
1. Optimal Decision Tree: Finding an optimal decision tree is an NP-complete
problem. Hence, decision tree induction algorithms employ a heuristic based
approach to search for the best in a large search space. Majority of the algorithms
follow a greedy, top-down recursive divide-and-conquer strategy to build
decision trees.
2. Missing data and noise: Decision tree induction algorithms are quite robust to
the data set with missing values and presence of noise. However, proper data pre-
processing can be followed to nullify these discrepancies.
3. Redundant Attributes: The presence of redundant attributes does not adversely

affect the accuracy of decision trees. It is observed that if an attribute is chosen
for splitting, then another attribute which is redundant is unlikely to chosen for
splitting.
4. Computational complexity: Decision tree induction algorithms are

computationally inexpensive, in particular, when the sizes of training sets are
large, Moreover, once a decision tree is known, classifying a test record is
extremely fast, with a worst-case time complexity of O(d), where d is the
maximum depth of the tree.
5. Data Fragmentation Problem: Since the decision tree induction algorithms
employ a top-down, recursive partitioning approach, the number of tuples
becomes smaller as we traverse down the tree. At a time, the number of
tuples may be too small to make a decision about the class representation,
such a problem is known as the data fragmentation. To deal with this
problem, further splitting can be stopped when the number of records falls
below a certain threshold.
6. Tree Pruning: A sub-tree can replicate two or more times in a decision tree
(see figure below). This makes a decision tree unambiguous to classify a test
record. To avoid such a sub-tree replication problem, all sub-trees except one
can be pruned from the tree.
A
C B
1 0
D C
0 1 D 1
0 1
7. Decision tree equivalence: The different splitting criteria followed in
different decision tree induction algorithms have little effect on the
performance of the algorithms. This is because the different heuristic
measures (such as information gain (), Gini index () and Gain ratio () are
quite consistent with each other); also see the figure below.

Reference
The detail material related to this lecture can be found in
Data Mining: Concepts and Techniques, (3rd Edn.), Jiawei Han, Micheline Kamber,
Morgan Kaufmann, 2015.
Introduction to Data Mining, Pang-Ning Tan, Michael Steinbach, and Vipin Kumar,
Addison-Wesley, 2014

Any question?
You may post your question(s) at the Discussion Forum

maintained in the course Web page!

09 Decision Tree Induction

Uploaded by

Copyright:

Available Formats

09 Decision Tree Induction

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

09 Decision Tree Induction

Uploaded by

Copyright:

Available Formats

Lecture #9

Classification: Decision Tree Induction

Dr. Debasis Samanta

Use of Decision Tree to classify data

Basic algorithm to build Decision Tree

Decision Tree induction algorithms

CS 40003: Data Analytics 2

Example 9.1: Binary Decision Tree

CS 40003: Data Analytics 3

Example 9.2: Decision Tree with numeric data

CS 40003: Data Analytics 4

There is a special node called root node.

Edges of a node represent the outcome for a value of the node.

In a path, a node with same label is never repeated.

Decision tree is not unique, as different ordering of internal nodes can

CS 40003: Data Analytics 5

Edges are the values of attributes

External nodes are the outcome of classification

Such a classification is, in fact, made by posing questions starting from

CS 40003: Data Analytics 6

What are the class label of Dragon and Shark?

CS 40003: Data Analytics 7

Gila Monster cold scale no no no yes yes

CS 40003: Data Analytics 8

Once a decision tree is built, it is applied to any test to classify it.

CS 40003: Data Analytics 9

Given a database D = 1 , 2 , . . , , where denotes a tuple, which is

CS 40003: Data Analytics 10

Some of them may give inaccurate result

Two approaches are known

Modification of greedy strategy

CS 40003: Data Analytics 11

CS 40003: Data Analytics 12

Case: Binary attribute

CS 40003: Data Analytics 13

Binary splitting by grouping attribute values

CS 40003: Data Analytics 14

Muti-way split: It is same as in the case of nominal attribute

CS 40003: Data Analytics 15

Binary outcome: A >v or A v

Here, q should be decided a priori

For a numeric attribute, decision tree induction is a combinatorial

CS 40003: Data Analytics 16

Person Gender Height Class

CS 40003: Data Analytics 17

Further, for each ordering, we can choose different ways of splitting

Different instances are shown in the following.

Approach 1 : <Gender, Height>

CS 40003: Data Analytics 18

CS 40003: Data Analytics 19

CS 40003: Data Analytics 20

A1 A2 A3 A4 Class Is there any clue that enables to

CS 40003: Data Analytics 21

CS 40003: Data Analytics 22

More ordered Less ordered

CS 40003: Data Analytics 23

CS 40003: Data Analytics 24

Two sheets showing the tabulation of marks obtained in a course

Which tabulation of marks shows the good performance of the