Data Mining

Lecture # 7
Naïve Bayes Classifier

1
 Naïve Bayes Classifier

Thomas Bayes
1702 - 1761

We will start off with a visual intuition, before looking at the math…
 Background
 There are three methods to establish a classifier
a) Model a classification rule directly
Examples: k-NN, decision trees, perceptron, SVM
b) Model the probability of class memberships given input data
Example: multi-layered perceptron with the cross-entropy
cost
c) Make a probabilistic model of data within each class
Examples: naive Bayes, model based classifiers
 a) and b) are examples of discriminative classification
 c) is an example of generative classification
 b) and c) are both examples of probabilistic classification
Grasshoppers Katydids

10
9
8
Antenna Length

7
6
5
4
3
2
1

1 2 3 4 5 6 7 8 9 10
Abdomen Length
Remember this example? Let’s
get lots more data…
With a lot of data, we can build a histogram.
Let us just build one for “Antenna Length” for
now…
10
9
8
Antenna Length

7
6
5
4
3
2
1

1 2 3 4 5 6 7 8 9 10

Katydids
Grasshoppers
We can leave the
histograms as they are, or
we can summarize them
with two normal
distributions.

Let us use two normal

distributions for ease of
visualization in the following
slides…
• We want to classify an insect we have found. Its antennae are 3 units
long. How can we classify it?

• We can just ask ourselves, give the distributions of antennae lengths

we have seen, is it more probable that our insect is a Grasshopper or
a Katydid.
• There is a formal way to discuss the most probable classification…
p(cj | d) = probability of class cj, given that we have observed d

Antennae length is 3
p(cj | d) = probability of class cj, given that we have observed d

P(Grasshopper | 3 ) = 10 / (10 + 2) = 0.833

P(Katydid | 3 ) = 2 / (10 + 2)= 0.166

10

Antennae length is 3
p(cj | d) = probability of class cj, given that we have observed d

P(Grasshopper | 7 ) = 3 / (3 + 9)= 0.250

P(Katydid | 7 ) = 9 / (3 + 9) = 0.750

Antennae length is 7
p(cj | d) = probability of class cj, given that we have observed d

P(Grasshopper | 5 ) = 6 / (6 + 6)= 0.500

P(Katydid | 5 ) = 6 / (6 + 6) = 0.500

6 6

Antennae length is 5
 Bayes Classifiers
That was a visual intuition for a simple case of the Bayes classifier, also called:

• Idiot Bayes
• Naïve Bayes
• Simple Bayes

We are about to see some of the mathematical formalisms, and more examples,
but keep in mind the basic idea.

Find out the probability of the previously unseen instance belonging to each
class, then simply pick the most probable class.
 Self Study
Concepts related to probability
 Probability Basics
• Prior, conditional and joint probability
– Prior probability:P(X)
– Conditional probability: P(X1 |X2 ), P(X2 |X1 )
– Joint probability:X  (X1 , X2 ), P(X)  P(X1 ,X2 )
– Relationship:P(X1 ,X2 )  P(X2 |X1 )P(X1 )  P(X1 |X2 )P(X2 )
– Independence:P(X |X )  P(X ), P(X |X )  P(X ), P(X ,X )  P(X )P(X )
2 1 2 1 2 1 1 2 1 2
• Bayesian Rule

P( X|C)P(C) Likelihood Prior

P(C |X)  Posterior 
P( X) Evidence
 Probabilistic Classification
• Establishing a probabilistic model for classification
– Discriminative model
P(C|X) C  c1 ,  , cL , X  (X1 ,  , Xn )
– Generative model
P(X|C) C  c1 ,  , cL , X  (X1 ,  , Xn )
• MAP classification rule
– MAP: Maximum A Posterior
– Assign x to c* if P(C  c * |X  x)  P(C  c |X  x) c  c * , c  c ,  , c
1 L
• Generative classification with the MAP rule
– Apply Bayesian rule to convert: P(X|C)P(C)
P(C |X)   P(X|C)P(C)
P( X)
 Naïve Bayes
• Bayes classification
P(C|X)  P(X|C)P(C)  P(X1 ,  , Xn |C)P(C)
Difficulty: learning the joint probabilityP(X ,  , X |C)
1 n

• Naïve Bayes classification

– Making the assumption that all input attributes are independent
P( X1 , X2 ,  , Xn |C )  P( X1 | X2 ,  , Xn ; C )P( X2 ,  , Xn |C )
 P( X1 |C )P( X2 ,  , Xn |C )
 P( X1 |C )P( X2 |C )    P( Xn |C )
– MAP classification rule

[ P( x1 |c * )    P( xn |c* )]P(c* )  [ P( x1 |c)    P( xn |c)]P(c), c  c* , c  c1 ,  , cL

 Naïve Bayes
• Naïve Bayes Algorithm (for discrete input attributes)
– Learning Phase: Given a training set S,
For each target value of ci (ci  c1 ,  , c L )
Pˆ (C  ci )  estimate P(C  ci ) with examplesin S;
For every attribute value a jk of each attribute x j ( j  1,  , n; k  1,  , N j )
Pˆ ( X j  a jk |C  ci )  estimate P( X j  a jk |C  ci ) with examplesin S;
Output: conditional probability tables; forx N  L elements
j, j

– Test Phase: Given an unknown instance X  (a ,  , a ) ,

1 n
Look up tables to assign the label c* to X’ if

[ Pˆ ( a1 |c * )    Pˆ ( an |c * )]Pˆ (c * )  [ Pˆ ( a1 |c)    Pˆ ( an |c)]Pˆ (c), c  c * , c  c1 ,  , c L

 Example
• Example: Play Tennis
 Example
• Learning Phase
Outlook Play=Yes Play=No Temperatur Play=Yes Play=No
Sunny e
2/9 3/5
Hot 2/9 2/5
Overcast 4/9 0/5
Mild 4/9 2/5
Rain 3/9 2/5
Cool 3/9 1/5
Humidity Play=Ye Play=N Wind Play=Yes Play=No
s o Strong 3/9 3/5
High 3/9 4/5 Weak 6/9 2/5
Normal 6/9 1/5

P(Play=Yes) = 9/14 P(Play=No) = 5/14

 Example
• Test Phase
– Given a new instance,
x’=(Outlook=Sunny, Temperature=Cool, Humidity=High, Wind=Strong)
– Look up tables
P(Outlook=Sunny|Play=Yes) = 2/9 P(Outlook=Sunny|Play=No) = 3/5
P(Temperature=Cool|Play=Yes) = 3/9 P(Temperature=Cool|Play==No) = 1/5
P(Huminity=High|Play=Yes) = 3/9 P(Huminity=High|Play=No) = 4/5
P(Wind=Strong|Play=Yes) = 3/9 P(Wind=Strong|Play=No) = 3/5
P(Play=Yes) = 9/14 P(Play=No) = 5/14

– MAP rule
P(Yes|x’): [P(Sunny|Yes)P(Cool|Yes)P(High|Yes)P(Strong|Yes)]P(Play=Yes) =
0.0053
P(No|x’): [P(Sunny|No) P(Cool|No)P(High|No)P(Strong|No)]P(Play=No) =
0.0206

Given the fact P(Yes|x’) < P(No|x’), we label x’ to be “No”.

 Example
• Test Phase
– Given a new instance,
x’=(Outlook=Sunny, Temperature=Cool, Humidity=High, Wind=Strong)
– Look up tables
P(Outlook=Sunny|Play=Yes) = 2/9 P(Outlook=Sunny|Play=No) = 3/5
P(Temperature=Cool|Play=Yes) = 3/9 P(Temperature=Cool|Play==No) = 1/5
P(Huminity=High|Play=Yes) = 3/9 P(Huminity=High|Play=No) = 4/5
P(Wind=Strong|Play=Yes) = 3/9 P(Wind=Strong|Play=No) = 3/5
P(Play=Yes) = 9/14 P(Play=No) = 5/14

– MAP rule
P(Yes|x’): [P(Sunny|Yes)P(Cool|Yes)P(High|Yes)P(Strong|Yes)]P(Play=Yes) =
0.0053
P(No|x’): [P(Sunny|No) P(Cool|No)P(High|No)P(Strong|No)]P(Play=No) =
0.0206

Given the fact P(Yes|x’) < P(No|x’), we label x’ to be “No”.

Assume that we have two classes (Note: “Drew
c1 = male, and c2 = female. can be a male
or female
name”)
We have a person whose sex we do not
know, say “drew” or d. Drew Barrymore
Classifying drew as male or female is
equivalent to asking is it more probable
that drew is male or female, I.e which is
greater p(male | drew) or p(female |
drew)
Drew Carey
What is the probability of being
called “drew” given that you are a
male? What is the
probability of being a
p(male | drew) = p(drew | male ) p(male) male?
p(drew) What is the probability
of being named “drew”?
(actually irrelevant, since it is
that same for all classes)
 This is Officer . Is Officer Drew a Male
or Female?
Luckily, we have a small database
with names and sex.

We can use it to apply Bayes rule…

Name Sex
Drew Male
Officer Drew Claudia Female
Drew Female
Drew Female
p(cj | d) = p(d | cj ) p(cj) Alberto Male
p(d) Karin Female
Nina Female
Sergio Male
 Name Sex
Drew Male
Claudia Female
Drew Female
Drew Female
p(cj | d) = p(d | cj ) p(cj) Alberto Male
p(d) Karin Female
Officer Drew
Nina Female
Sergio Male
p(male | drew) = 1/3 * 3/8 = 0.125
3/8 3/8
Officer Drew is more
likely to be a Female.
p(female | drew) = 2/5 * 5/8 = 0.250
3/8 3/8
 Officer Drew IS a female!

Officer Drew
So far we have only considered Bayes
Classification when we have one attribute (the p(cj | d) = p(d | cj ) p(cj)
“antennae length”, or the “name”). But we may
have many features.
p(d)
How do we use all the features?

Name Over 170CM Eye Hair length Sex

Drew No Blue Short Male
Claudia Yes Brown Long Female
Drew No Blue Long Female
Drew No Blue Long Female
Alberto Yes Brown Short Male
Karin No Blue Long Female
Nina Yes Brown Short Female
Sergio Yes Blue Long Male
 • To simplify the task, naïve Bayesian classifiers assume
attributes have independent distributions, and thereby
estimate

p(d|cj) = p(d1|cj) * p(d2|cj) * ….* p(dn|cj)

The probability of class cj
generating instance d,
equals….

The probability of class

cj generating the
observed value for
feature 1, multiplied by..
The probability of class
cj generating the
observed value for
feature 2, multiplied by..
• To simplify the task, naïve Bayesian classifiers
assume attributes have independent distributions,
and thereby estimate
p(d|cj) = p(d1|cj) * p(d2|cj) * ….* p(dn|cj)

p(officer drew|cj) = p(over_170cm = yes|cj) * p(eye =blue|cj) * ….

Officer Drew is
blue-eyed,
over 170cm tall, p(officer drew| Female) = 2/5 * 3/5 * ….
and has long p(officer drew| Male) = 2/3 * 2/3 * ….
hair
The Naive Bayes classifiers is
often represented as this type of
graph… cj
Note the direction of the arrows,
which state that each class
causes certain features, with a
certain probability

p(d1|cj) p(d2|cj) … p(dn|cj)

Naïve Bayes is fast cj
and space efficient

We can look up all the

probabilities with a single scan of
the database and store them in a
(small) table…
p(d1|cj) p(d2|cj) … p(dn|cj)

Sex Over190cm Sex Long Hair Sex

Male Yes 0.15 Male Yes 0.05 Male
No 0.85 No 0.95
Female Yes 0.01 Female Yes 0.70 Female
No 0.99 No 0.30
Naïve Bayes is NOT sensitive to irrelevant features...

Suppose we are trying to classify a persons sex

based on several features, including eye color. (Of
course, eye color is completely irrelevant to a
persons gender)
p(Jessica |cj) = p(eye = brown|cj) * p( wears_dress = yes|cj) * ….
p(Jessica | Female) = 9,000/10,000 * 9,975/10,000 * ….
p(Jessica | Male) = 9,001/10,000 * 2/10,000 * ….

Almost the same!

However, this assumes that we have good enough estimates of the

probabilities, so the more data the better.
An obvious point. I have used a simple
two class problem, and two possible
values for each example, for my previous
cj
examples. However we can have an
arbitrary number of classes, or feature
values

p(d1|cj) p(d2|cj) … p(dn|cj)

Animal Color
Animal Mass >10kg Cat Black 0.33 Animal
Cat Yes 0.15 White 0.23 Cat
No 0.85 Brown 0.44
Dog Yes 0.91 Dog
Dog Black 0.97
No 0.09 White 0.03
Pig
Pig Yes 0.99 Brown 0.90
No 0.01 Pig Black 0.04
White 0.01
Brown 0.95
Problem! Naïve Bayesian
p(d|cj) Classifier
Naïve Bayes assumes
independence of
features…

p(d1|cj) p(d2|cj) p(dn|cj)

Sex Over 6 Sex Over 200

foot pounds
Male Yes 0.15 Male Yes 0.11
No 0.85 No 0.80
Female Yes 0.01 Female Yes 0.05
No 0.99 No 0.95
Solution Naïve Bayesian
p(d|cj) Classifier
Consider the relationships
between attributes…

p(d1|cj) p(d2|cj) p(dn|cj)

Sex Over 6 Sex Over 200 pounds

foot
Male Yes and Over 6 foot 0.11
Male Yes 0.15
No and Over 6 foot 0.59
No 0.85
Yes and NOT Over 6 foot 0.05
Female Yes 0.01
No and NOT Over 6 foot 0.35
No 0.99
Female Yes and Over 6 foot 0.01
Solution Naïve Bayesian
p(d|cj) Classifier
Consider the relationships
between attributes…

p(d1|cj) p(d2|cj) p(dn|cj)

But how do we find the set of connecting arcs??

The Naïve Bayesian Classifier
has a quadratic decision boundary
10
9
8
7
6
5
4
3
2
1

1 2 3 4 5 6 7 8 9 10
 Relevant Issues
• Violation of Independence Assumption
– For many real world tasks,P(X1 ,  , Xn |C)  P(X1 |C)    P(Xn |C)
– Nevertheless, naïve Bayes works surprisingly well anyway!
• Zero conditional probability Problem
– If no example contains the attribute valueX  a , Pˆ ( X  a |C  c )  0
j jk j jk i
– In this circumstance, Pˆ ( x |c )    Pˆ (a |c )    Pˆ ( x |c )  0 during test
1 i jk i n i
– For a remedy, conditional probabilities estimated with
n  mp
Pˆ ( X j  a jk |C  ci )  c
nm
nc : numberof training examplesfor which X j  a jk and C  ci
n : numberof training examplesfor which C  ci
p : prior estimate(usually, p  1 / t for t possible values of X j )
m : weight to prior (numberof " virtual" examples, m  1)
 Estimating Probabilities
• Normally, probabilities are estimated based on observed
frequencies in the training data.
• If D contains nk examples in category yk, and nijk of these nk
examples have the jth value for feature Xi, xij, then:
nijk
P( X i  xij | Y  yk ) 
nk
• However, estimating such probabilities from small training
sets is error-prone.

– If due only to chance, a rare feature, Xi, is always false in the

training data, yk :P(Xi=true | Y=yk) = 0.
– If Xi=true then occurs in a test example, X, the result is that yk:
P(X | Y=yk) = 0 and yk: P(Y=yk | X) = 0
9/16/2013 4747
 Probability Estimation Example
Probability positive negative
Ex Size Color Shape Category
P(Y) 0.5 0.5
1 small red circle positive P(small | Y) 0.5 0.5
P(medium | Y) 0.0 0.0
2 large red circle positive
P(large | Y) 0.5 0.5
3 small red triangle negitive P(red | Y) 1.0 0.5
P(blue | Y) 0.0 0.5
4 large blue circle negitive
P(green | Y) 0.0 0.0
P(square | Y) 0.0 0.0

Test Instance X: P(triangle | Y) 0.0 0.5

<medium, red, circle> P(circle | Y) 1.0 0.5

P(positive | X) = 0.5 * 0.0 * 1.0 * 1.0 / P(X) = 0

P(negative
9/16/2013
| X) = 0.5 * 0.0 * 0.5 * 0.5 / P(X) = 0 4848
 Naïve Bayes Example
Probability positive negative
P(Y) 0.5 0.5
P(small | Y) 0.4 0.4
P(medium | Y) 0.1 0.2
P(large | Y) 0.5 0.4 Test Instance:
P(red | Y) 0.9 0.3 <medium ,red, circle>
P(blue | Y) 0.05 0.3
P(green | Y) 0.05 0.4
P(square | Y) 0.05 0.4
P(triangle | Y) 0.05 0.3
P(circle | Y) 0.9 0.3

9/16/2013 4949
 Naïve Bayes Example
Probability positive negative
P(Y) 0.5 0.5
P(medium | Y) 0.1 0.2
P(red | Y) 0.9 0.3 Test Instance:
<medium ,red, circle>
P(circle | Y) 0.9 0.3

P(positive | X) = P(positive)*P(medium | positive)*P(red | positive)*P(circle | positive) / P(X)

0.5 * 0.1 * 0.9 * 0.9
= 0.0405 / P(X) = 0.0405 / 0.0495 = 0.8181
P(negative | X) = P(negative)*P(medium | negative)*P(red | negative)*P(circle | negative) / P(X)
0.5 * 0.2 * 0.3 * 0.3
= 0.009 / P(X) = 0.009 / 0.0495 = 0.1818
P(positive | X) + P(negative | X) = 0.0405 / P(X) + 0.009 / P(X) = 1

P(X) = (0.0405 + 0.009) = 0.0495

9/16/2013 5050
 Smoothing
• To account for estimation from small samples,
probability estimates are adjusted or smoothed.
• Laplace smoothing using an m-estimate assumes
that each feature is given a prior probability, p, that
is assumed to have been previously observed in a
“virtual” sample of size m.
nijk  mp
P( X i  xij | Y  yk ) 
nk  m

• For binary features, p is simply assumed to be 0.5.

9/16/2013 5151
 Laplace Smothing Example
• Assume training set contains 10 positive examples:
– 4: small
– 0: medium
– 6: large
• Estimate parameters as follows (if m=1, p=1/3)
– P(small | positive) = (4 + 1/3) / (10 + 1) = 0.394
– P(medium | positive) = (0 + 1/3) / (10 + 1) = 0.03
– P(large | positive) = (6 + 1/3) / (10 + 1) = 0.576
– P(small or medium or large | positive) = 1.0

9/16/2013 5252
 Numerical Stability
• It is often the case that machine learning algorithms need to
work with very small numbers
– Imagine computing the probability of 2000 independent
coin flips
– MATLAB thinks that (.5)2000=0

9/16/2013 53
 Stochastic Language Models
• Models probability of generating strings (each
word in turn) in the language (commonly all
strings over ∑). E.g., unigram model
Model M

0.2 the
the guy likes the fruit
0.1 a
0.01 guy 0.2 0.01 0.02 0.2 0.01

0.01 fruit
0.03 said multiply
0.02 likes
P(s | M) = 0.00000008
…
9/16/2013 54
13.2.1
 Numerical Stability
• Instead of comparing P(Y=5|X1,…,Xn) with P(Y=6|X1,…,Xn),
– Compare their logarithms

9/16/2013 55
 Underflow Prevention
• Multiplying lots of probabilities, which are
between 0 and 1 by definition, can result in
floating-point underflow.
• Since log(xy) = log(x) + log(y), it is better to
perform all computations by summing logs of
probabilities rather than multiplying
probabilities.
• Class with highest final un-normalized log
probability score is still the most probable.
9/16/2013 5656
 Relevant Issues
• Continuous-valued Input Attributes
– Numberless values for an attribute
– Conditional probability modeled with the normal distribution
1  ( X j   ji )2 
Pˆ ( X j |C  ci )  exp  
2  ji  2 ji 
2

 ji : mean(avearage)of attribute values X j of examples for which C  ci
 ji : standarddeviation of attribute values X j of examples for which C  ci

– Learning Phase: for X  (X ,  , X ), C  c ,  , c

1 n 1 L
Output: n L normal distributions andP(C  c ) i  1,  , L
i
– Test Phase:
for X  (X1 ,  , Xn )
• Calculate conditional probabilities with all the normal distributions
• Apply the MAP rule to make a decision
 Data with Numeric Attributes
foot
sex height (feet) weight (lbs)
size(inches)
male 6 180 12
male 5.92 (5'11") 190 11
male 5.58 (5'7") 170 12
male 5.92 (5'11") 165 10
female 5 100 6
female 5.5 (5'6") 150 8
female 5.42 (5'5") 130 7
female 5.75 (5'9") 150 9

mean variance mean variance mean variance

sex
(height) (height) (weight) (weight) (foot size) (foot size)
3.5033e- 1.2292e+0 9.1667e-
male 5.855 176.25 11.25
02 2 01
9.7225e- 5.5833e+0 1.6667e+0
female 5.4175 132.5 7.5
02 2 0
Data with Numeric Attributes

foot
sex height (feet) weight (lbs)
size(inches)
sample 6 130 8
Since posterior numerator is greater in the female case, we predict the
sample is female.
Advantages/Disadvantages of Naïve Bayes
• Advantages:
– Fast to train (single scan). Fast to classify
– Not sensitive to irrelevant features
– Handles real and discrete data
– Handles streaming data well
• Disadvantages:
– Assumes independence of features
 Conclusions
 Naïve Bayes based on the independence assumption
 Training is very easy and fast; just requiring considering each
attribute in each class separately
 Test is straightforward; just looking up tables or calculating
conditional probabilities with normal distributions
 A popular generative model
 Performance competitive to most of state-of-the-art
classifiers even in presence of violating independence
assumption
 Many successful applications, e.g., spam mail filtering
 Apart from classification, naïve Bayes can do more…
 Conclusions
• Naïve Bayes is:
– Really easy to implement and often works well
– Often a good first thing to try
– Commonly used as a “punching bag” for smarter
algorithms

9/16/2013 64
 Acknowledgements
 Introduction to Machine Learning, Alphaydin
 Statistical Pattern Recognition: A Review – A.K Jain et al., PAMI (22) 2000
Material in these slides has been taken from, the following resources

 Pattern Recognition and Analysis Course – A.K. Jain, MSU

 Pattern Classification” by Duda et al., John Wiley & Sons.
 http://www.doc.ic.ac.uk/~sgc/teaching/pre2012/v231/lecture13.html
 http://en.wikipedia.org/wiki/Naive_Bayes_classifier
 Some Material adopted from Dr. Adam Prugel-Bennett

65