MATH 230

Introduction to Probability
Theory

Some Discrete Probability

Distributions

Copyright © 2010 Pearson Addison-Wesley. All rights reserved.

TEDU
Chapter 5 - Some Discrete
Probability Distributions
5.1 Introduction and Motivation
5.2 Binomial and Multinomial Distributions
5.3 Hypergeometric Distribution
5.4 Negative Binomial and Geometric Distributions
5.5 Poisson Distribution and the Poisson Process
Introduction and Motivation

 No matter whether a discrete probability distribution is

represented graphically, or by means of a formula, the
behavior of a random variable is described.
 Often, the observations generated by different statistical
experiments have the same general type of behavior.
 Consequently, discrete random variables associated with these
experiments can be described by essentially the same
probability distribution and therefore can be represented by a
single formula.
 In this chapter, we present these commonly used distributions
with various examples.
Copyright © 2010 Pearson Addison-Wesley. All rights reserved.
Bernoulli Distribution

The Bernoulli Process

Strictly speaking, the Bernoulli process must possess the following

properties:

1. The experiment consists of repeated trials.

2. Each trial results in an outcome that may be classified as a
success or a failure.
3. The probability of success, denoted by p, remains constant
from trial to trial.
4. The repeated trials are independent.
Bernoulli Distribution
Bernoulli Distribution

Mean and Variance

Binomial Distribution

Suppose we conduct n independent Bernoulli trials,

e.g., toss a coin n times.

 Each trial results in either a success with probability p

or a failure with probability 1-p.
 Then, the total number of successes in n trials is a binomial
random variable.
More formally,
Binomial Distribution

Example. Toss a coin 10 times. What is the probability

of getting three heads if P(H) = p = 0.3?

n = 10
X: the number of heads in 10 trials,
Binomial Distribution
Binomial Distribution
 Binomial Distribution

Where Does the Name Binomial Come From?

Binomial Distribution
Binomial Distribution

Mean and Variance

Using Bernoulli trials and properties of mean and variance,

Theorem 5.1

Copyright © 2010 Pearson Addison-Wesley. All rights reserved.

Binomial Distribution

Example 1. An examination consists of 10 multi-choice questions,

in each of which a candidate has to deduce which one of five
suggested answers is correct. A completely unprepared student
may be assumed to guess each answer completely randomly.

a) What is the probability that this student gets 8 or more

questions correct?
b) What is the expected value and standard deviation of the
number of correct answers?

Copyright © 2010 Pearson Addison-Wesley. All rights reserved.

Binomial Distribution

Let X be a random variable ‘number of answers guessed

correctly’ then for each question (i.e. trial) the probability of a
‘success’ =1/5 =0.2.
It is clear that X follows a binomial distribution with n=10 and
p=0.2.
P  X  8   P  X  8   P  X  9   P  X  10 
10  10  10 
    0.2   0.8      0.2   0.8      0.2   0.8 
8 2 9 1 10 0

8 9 10 

 0.000078
E  X   np  10*0.2  2
VAR  X   np 1  p   10*0.2*0.8  1.6
Copyright © 2010 Pearson Addison-Wesley. All rights reserved.
Binomial Distribution

Example 2. There are three urns, each containing 10 balls

numbered zero through nine. A person picks one ball from
each urn and wins if all three balls have the same number.
100 people play the game independently.

a) What is the probability that at least two people win the

game?
b) What is the expected value and standard deviation of the
number of winners?

Copyright © 2010 Pearson Addison-Wesley. All rights reserved.

Binomial Distribution

Let X be a random variable ‘number of winners’ then for each

game (i.e. trial) the probability of a ‘success’=(1)(0.1)(0.1)=0.01.
It is clear that X follows a binomial distribution with n=100 and
p=0.01.

P  X  2   1  P  X  2   1   P  X  0   P  X  1 
100  100  99 
 1     0.01  0.99      0.01  0.99    0.264
0 100 1

 0   1  

E  X   np  100*0.01  1
 2  VAR  X   np 1  p   100*0.01*0.99  0.99
  0.995
Copyright © 2010 Pearson Addison-Wesley. All rights reserved.
Hypergeometric Distribution

 The types of applications for the hypergeometric are very

similar to those for the binomial distribution.
 But in the case of the binomial distribution, independence
among trials is required.
 if binomial distribution is applied to, say, sampling from a lot
of items (deck of cards, batch of production items), the
sampling must be done with replacement of each item after
it is observed.
 On the other hand, the hypergeometric distribution does not
require independence and is based on sampling done
without replacement.
Hypergeometric Distribution

Hypergeometric Experiment
1. A random sample of size n is selected without replacement from
N items.
2. Of the N items, k may be classified as successes and N − k are
classified as failures.
The number X of successes of a hypergeometric experiment is
called a hypergeometric random variable.
Accordingly, the probability distribution of the hypergeometric
variable is called the hypergeometric distribution, and its values
are denoted by h(x; N, n, k), since they depend on the number of
successes k in the set N from which we select n items.
Hypergeometric Distribution

Example: Lots of 40 components each are deemed

unacceptable if they contain 3 or more defectives. The
procedure for sampling a lot is to select 5 components at
random and to reject the lot if a defective is found.

a. What is the probability that exactly 1 defective is found in

the sample if there are 3 defectives in the entire lot?

b. Comment on the utility of this plan.

Copyright © 2010 Pearson Addison-Wesley. All rights reserved.

 Hypergeometric Distribution

Solution:
a. Using the hypergeometric distribution with n = 5, N = 40, k = 3,
and x = 1, we find the probability of obtaining 1 defective to be

b. This plan is not desirable since it accepts a bad lot 66% of the
time since it detects a bad lot (3 defectives) only about 34% of the
time.
 3  37 
  
Accept the lot when X=0 where h  0, 40,5,3      0.66
0 5
 40 
 
5
Copyright © 2010 Pearson Addison-Wesley. All rights reserved.
Hypergeometric Distribution

Copyright © 2010 Pearson Addison-Wesley. All rights reserved.

Hypergeometric Distribution

Theorem 5.2

Recall Example:
Find the mean and variance for the number of defective items
found in the sample when 5 components are selected at random
from a lot of 40 components of which 3 are defective.
Geometric Distribution

 Suppose we again conduct a series of independent Bernoulli

trials where P(success) = p and P(failure) = 1-p.
 Then, the number of trials until first success (including the
successful trial) is a geometric random variable.
 P(X=n)=P(Xi=0 for i=1,…,n-1, and Xn=1) where Xi’s are
Bernoulli r.v.s.
 Geometric random variable X can take on values 1, 2, ...
 Hence, it is an example of a random variable whose domain
contains countably infinite number of values.
Geometric Distribution

Example. For a certain manufacturing process, it is known that, on

the average, 1 in every 100 items is defective. What is the
probability that the fifth item inspected is the first defective item
found?

Solution : Random Variable

X: the number of items inspected until the first defective item is
found.
p=0.01 (probability that the inspected item will be defective.)

P(X=5) = (0.99)4(0.01)= 0.0096.

 Geometric Distribution

Example. A professional thinks that, in any job she applies, her chances
of getting a job offer is 0.4. What is the probability that she will have
to apply at most four job openings before she gets an offer?
Solution. p = probability of getting a job offer = 0.4
X: the number of applications until the first offer, x =1, 2, …
P(X ≤ 4) = p(1) + p(2) + p(3) + p(4) where
p(1) = 0.4
p(2) = 0.6⋅0.4 = 0.24
p(3) = 0.6⋅0.6⋅0.4 = 0.144
p(4) = 0.6⋅0.6⋅0.6⋅0.4 = 0.0864
P(X ≤ 4) = 0.4 + 0.24 + 0.144 + 0.0864 = 0.8704
Geometric Distribution

Copyright © 2010 Pearson Addison-Wesley. All rights reserved.

Geometric Distribution

Copyright © 2010 Pearson Addison-Wesley. All rights reserved.

Geometric Distribution

Theorem 5.3

Copyright © 2010 Pearson Addison-Wesley. All rights reserved.

 Geometric Distribution

Example. A professional thinks that, in any job she applies, her

chances of getting a job offer is 0.4. What is the expected
number and standard deviation of jobs the professional has to
apply before she gets a job offer?
Random Variable: X: the number of applications until the first offer

X ~ Geo  0.4 
1 1
EX     2.5
p 0.4
1  p 0.6
  2  2  3.75
2

p 0.4
  3.75  1.936
 “Memoryless” Property
of Geometric Distribution

Example: Let p=0.2 be the probability of a success in a trial. Probability

of getting the first success at the end of a series of 3 trials:
P( X  3)   0.8  0.2 
2

Probability of getting the first success at the 4th trial given that
we have failure in the first trial:
P( X  4 )  0.8  0.2 
3
P( X  4 and X  1)
P( X  4 | X  1)      0.8   0.2   P( X  3)
2

P  X  1 1  P  X  1 1   0.2 

P( X  5 | X  2)   0.8   0.2   P( X  3)
2
Similarly:
P( X  6 | X  3)   0.8   0.2   P( X  3)
2

...
That is; P( X  s  3 | X  s )   0.8   0.2   P( X  3)
2
“Memoryless” Property
of Geometric Distribution

 This means that, probability of getting the first success at the

end of a series of x trials, which we will start now, is
independent of the number of trials (failures) we had in the
past. Hence, the distribution is memoryless.
 Geometric distribution is the only discrete distribution having
this property.
“Memoryless” Property
of Geometric Distribution
Negative Binomial Distribution
(Pascal Distribution)

 We again conduct a series of independent Bernoulli trials

where P(success) = p and P(failure) =1-p.
 Then, the number of trials until rth succes (including the
successful trial) is a Pascal random variable.
 Pascal distribution reduces to geometric distribution for r = 1.
 Pascal distribution is also known as negative binomial
distribution.
Negative Binomial Distribution
(Pascal Distribution)

Example: In an NBA championship series, the team that wins

four games out of seven is the winner. Suppose that teams A and
B face each other in the championship games and that team A has
probability 0.55 of winning a game over team B.
What is the probability that team A will win the series in 6th
game?

 5
P( X  6)     0.55   0.45   0.55   0.1853
3 2

 3
Negative Binomial Distribution
(Pascal Distribution)
Negative Binomial Distribution
(Pascal Distribution)
Negative Binomial Distribution
(Pascal Distribution)
 Poisson Distribution

 Experiments yielding numerical values of a random variable X, the

number of outcomes occurring during a given time interval or in a
specified region, are called Poisson experiments.
 Poisson distribution is used to model occurrence of rare events.

For example, a Poisson experiment can generate observations for the

random variable X representing:
 the number of telephone calls received per hour by an office,
 the number of days school is closed due to snow during the winter.
 the number of field mice per acre, the number of bacteria in a given
culture, or
the number of typing errors per page.
λ : the average number of occurences per unit time or region,
Poisson Distribution
Properties of the Poisson Process
1. The number of outcomes occurring in one time interval or specified
region of space is independent of the number that occur in any other
disjoint time interval or region.
2. The probability that a single outcome will occur during a very short
time interval or in a small region is proportional to the length of the time
interval or the size of the region and does not depend on the number of
outcomes occurring outside this time interval or region.
3. The probability that more than one outcome will occur in such a
short time interval or fall in such a small region is negligible.
The number X of outcomes occurring during a Poisson experiment is
called a Poisson random variable, and its probability distribution is
called the Poisson.
Poisson Distribution

Theorem 5.4

Copyright © 2010 Pearson Addison-Wesley. All rights reserved.

 Poisson Distribution
EXAMPLE: The number of customers arriving per hour at a certain
automobile service facility is assumed to follow a Poisson distribution at
the rate of 7 per hour.
(a) Compute the probability that exactly 10 customers will arrive in a 2-
hour period.
X: number of customers that arrive during a 2-hour period, X ~ Poisson (14)
e141410
P  X  10    0.066
10!

(b) Compute the probability that at least 3 customers will arrive in a 2-

hour period. P X  3  1  P X  3  1  2 e1414x  0.99
    
x 0 x!
(c) What is the mean and average number of arrivals during a 2-hour
period? E(X) =Var(X)= λt =14
Nature of the Poisson Probability
Function
Like so many discrete and continuous distributions, the form of the Poisson
distribution becomes more and more symmetric, even bell-shaped, as
the mean grows large.

Figure 5.1 Poisson density functions for different means

Approximation of Binomial Distribution
by a Poisson Distribution

 Poisson distribution is related to the binomial distribution.

 The independence among Bernoulli trials in the binomial case is
consistent with principle of Poisson process.
 Poisson distribution is used to model occurrence of rare events.
 If n is large and p is close to 0, the Poisson can be used, with μ=
np, to approximate binomial probabilities.
 It is a limiting distribution for binomial distribution as n∞ and
p0 such that np remains constant.
Approximation of Binomial Distribution
by a Poisson Distribution

Example: In a certain industrial facility, accidents occur infrequently.

It is known that the probability of an accident on any given day is
0.005 and accidents are independent of each other.
(a) What is the probability that in any given period of 400 days
there will be an accident on one day?
(b) What is the probability that there are at most three days with an
accident in any given period of 400 days?
Let X be a binomial random variable with n = 400 and p = 0.005.
Thus, np = 2. Using the Poisson approximation,
 400  e2 21
a. P  X  1     0.005   0.995   0.271 
1 399

 1  1!
3
e2 2 x
b. P  X  3    0.857
x 0 x!
Approximation of Binomial Distribution
by a Poisson Distribution

Theorem 5.5

Copyright © 2010 Pearson Addison-Wesley. All rights reserved.

Approximation of Binomial Distribution
by a Poisson Distribution
Approximation of Binomial Distribution
by a Poisson Distribution
Approximation of Binomial Distribution
by a Poisson Distribution
Approximation of Binomial Distribution
by a Poisson Distribution