CHAPTER 4

SPECIAL DISCRETE
PROBABILITY DISTRIBUTIONS
(PART 3)
1

I [email protected]
2 SUBTOPICS
❑Special Discrete Probability Distributions
➢ Uniform
➢ Binomial
➢ Poisson
❑Poisson Approximation to Binomial Distribution
 In this lesson, you will learn:
3

1) The properties/conditions needed to model

each discrete random variable
2) For each distribution, you need to know how to:
a) calculate the probability
b) the mean and variance/standard deviation
3) Use the Poisson distribution as an approximation
to the Binomial distribution
4
THE UNIFORM DISTRIBUTION

[email protected]
5
PROPERTIES of Discrete Uniform
Distribution
The discrete r. v. X assumes each of its value with an equal
probability.
Minimum value is fixed.
Maximum value is fixed.
All values in range are equally likely to occur.
 The Discrete Uniform Distribution
If the r. v. X assumes the values x1, x2, …, xk with equal
probabilities, then the discrete uniform distribution is
given by:
1 1
f ( x) = , x = x1 , x2 ,..., xk or P ( X = xr ) = , r = 1, 2, ..., k
k k
NOTE: k is the parameter.
The mean and variance of discrete uniform distribution are:

E ( X ) =  x. f ( x)
V ( X ) = E ( X 2 ) − [ E ( X )]2

06
 Example 1
7

When a light bulb is selected at random from a box that contains

a 40-watt bulb, a 60-watt bulb, a 75-watt bulb and a 100-watt
bulb, each element of the sample space S={40, 60, 75, 100} occurs
with probability ¼. Therefore we have the uniform distribution with

1
f ( x) = , x = 40, 60, 75, 100
4
x 40 60 75 100
f (x)=P(X=x) 1/4 1/4 1/4 1/4

• Find the mean and variance of the random variable X.

• Find P(X < ).

09
 Example 1 Solution
8
x 40 60 75 100 E ( X ) =  x. f ( x)
f (x)=P(X=x) 1/4 1/4 1/4 1/4 V ( X ) = E ( X 2 ) − [ E ( X )]2
1
f ( x) =
4
1  1
E ( X ) =  x. f ( x) =  x  = (40 + 60 + 75 + 100) = 68.75
4  4
1 1
E ( X 2 ) =  x 2 f ( x) =  x 2
= (40 2 + ... + 100 2 ) = 5206.25
4 4
Var ( X ) = E ( X 2 ) − [ E ( X )]2 = 5206.25 − (68.75) 2 = 479.69

P( X   ) = P( X  68.75)
= P( X = 40) + P( X = 60)
10 = 0.25 + 0.25 = 0.5
 THE BERNOULLI AND
9
BINOMIAL DISTRIBUTIONS

[email protected]
10 PROPERTIES of Bernoulli EXPERIMENT
1. Each trial has only two possible outcomes that may be labeled
as a success or
a failure.
2. For a Bernoulli trial (experiment is done only once), the
probability of success
is denoted as p and failure as q = 1 – p.
Examples:
Testing of items as they come off an assembly line where each test or trial may
indicate a defective or nondefective item.
tossing a coin (head or tail), gender (male or female), win or lose, true or false
question, yes/no, etc.

Note: Classification of “success” and “failure” does not imply that one outcome is
good and the other is bad.
11
THE Bernoulli DISTRIBUTION
The probability distribution for Bernoulli is

f ( x) = p x (1 − p )1− x = p x q1− x , x = 0 or 1

The mean and variance are:

 = E( X ) = p
 2 = V ( X ) = p(1 − p) = pq
Binomial Distribution
The binomial distribution is just n independent Bernoullis
added up.
It is the number of “successes” in n trials.
If Z1, Z2, …, Zn are Bernoulli, then X is binomial:
X = Z1 + Z 2 +  + Z n

We are interested in X, the number of successes observed

during the n trials, for x = 0, 1, 2, …, n

12
13
Properties of Binomial Experiment

1. There are n identical trials.

2. Each trial has only two possible outcomes
(a success or a failure).
3. The probability of success is p and remains constant
from trial to trial.
4. The repeated trials are independent.
5. The random variable of interest is X, the number of
successes observed during n trials.
14 The Binomial Distribution
Then the probability distribution of the binomial random variable
X, the number of successes in n independent trials, is
n− x
b( x; n, p) = P( X = x)= C x p q
n x
, x = 0, 1, 2, ..., n
where
n = total number of trials
n n!
p = probability of success
n
C x =   =
q = 1 – p = probability of failure  x  x!(n − x)!
x = number of successes in n trials
n - x = number of failures in n trials

The mean and variance for  = E ( X ) = np

Binomial distribution are:  2 = V ( X ) = np(1 − p) = npq
15
Example 2
At Giant Supermarket, 60% of customers pay by
credit card. Find the probability that in a randomly
selected sample of 10 customers,
(a)exactly two pay by credit card
(b)more than 7 pay by credit card.
16
Example 2 Solution
X = no of customers in a sample of 10 who pay by
credit card
p= P(pay by credit card) = 0.6, q = 1-p = 0.4
n= 10, assuming independence, so X ~B(10, 0.6).
(a) P( X = 2)=10 C p 2 q 8 = 45  0.6 2  0.48 = 0.0106
2

(b) P( X  7) = P( X = 8) + P( X = 9) + P( X = 10)
=10 C8 p 8 q 2 +10C9 p 9 q1 +10C10 p10 q 0
= (45  0.68  0.4 2 ) + (10  0.69  0.41 ) + 0.610 = 0.1673
17
Example 3
Over a long period of time, it has been observed that a given
rifleman can hit a target on a single trial with probability equal to
0.8. Suppose he fires four shots at the target. What is the
probability that he will hit the target:
(a) exactly two times?
(b) at least once?
Example 3 Solution
18

X= number of shots that hit the target

n = 4 ; p= P(hit the target)=0.8, q =1 – p = 0.2
X ~ Bin(4, 0.8)

(a) P ( X = 2)= 4C  0.82  0.2 2 = 0.1536

2

(b) P( X  1) = 1 − P( X  1)
= 1 − P( X = 0)
= 1 −( 4C0  0.80  0.2 4 )
= 0.9984
Using The Cumulative Binomial Probabilities Table
19

The cumulative binomial

probabilities table provided
gives the values of which is
the probability of
n
n x
P( X  r ) =    p (1 − p) n − x
x=r  x 

obtaining r or more
successes in n independent
trials.
Refer Statistical Table 1
20 Example 4
The probability that a patient recovers from a rare blood disease is
0.4. If 10 people are known to have contracted this disease.
Find the probability that:
(a) At least 7 survive.
(b) More than 7 survive.
(c) From 3 to 5 survive.
(d) Exactly 5 survive.
 Example 4 Solution
21

X = the number of people that survive

p = P(patient survives)=0.4, q=1 – p =0.6
n = 10, X~Bin(10, 0.4)
Using Table 1,
(a) P(X ≥ 7) = 0.0548

(b) P(X > 7) = P(X ≥ 8)

= 0.0123

Note: Copy the values

exactly from the table
 Example 4 Solution cont.
22

X = the number of people that survive

p = P(patient survives)=0.4, q=1 – p =0.6
n = 10, X~Bin(10, 0.4)

Using Table 1, Note: Alternatively, you can

just use the formula

(c) P(3  X  5) = P(X ≥ 3) – P(X ≥ 6)

= 0.8327 – 0.1662 = 0.6665

(d) P(X = 5) = P(X ≥ 5) – P(X ≥ 6)

= 0.3669 – 0.1662 = 0.2007
 THE POISSON
23
DISTRIBUTION

[email protected]
 THE Poisson Distribution
24
Consider the discrete random variable X:
➢ The number of emergency calls received by an
ambulance in an hour
➢ The number of vehicles approaching the Ayer Keroh Toll in
a five-minute interval
➢ The number of flaws in a meter length of material
➢ The number of field mice per acre
➢ The number of bacteria in a given culture
➢ The number of typing errors per page
Assuming that each occurs randomly, above are
examples of variables that can be modeled using
Poisson distribution.
25
PROPERTIES of Poisson EXPERIMENT
Events occur at random in a given interval of time or
space
The r. v. is the number of events that occur in an
interval of a given size.
Events occur independently.
λ, the mean number of occurrences in the given
interval is known and finite.
26
The Poisson Distribution
The probability distribution of the Poisson random variable X,
representing the number of outcomes occurring in a given time
interval or specified region, is

e −  ( ) x
p( x;  ) = P( X = x) = , x = 0, 1, 2, ...
x!

where ‘lambda’ is the average number of outcomes per unit

time, area, distance, or volume, and e=2.71828….
The mean and variance for Poisson distribution are:
 = E( X ) = 
 2 = V (X ) = 
27
Example 5
The number of cracks in a ceramic tile has Poisson distribution
with mean 2.4.
a) What is the probability that a ceramic tile has no crack?
b) What is the probability that a ceramic tile has four or more
cracks?
c) If five ceramic tiles are checked, what is the probability that
there are at least five cracks on them?
 Example 5 Solution
28

Let X = number of cracks in a ceramic tile (for 1 tile)

λ=2.4, X~Poi(2.4)

−2.4 0
(a) P( X = 0) = e 2 . 4
= 0.0907
0!
 e − 2.4 2.4 0 e − 2.4 2.41 e − 2.4 2.4 2 e − 2.4 2.4 3 
(b) P( X  4) = 1 −  + + +  = 0.2213
 0! 1! 2! 3! 
X ~ Po ( new = 5  2.4 = 12)
(c)
P( X  5) = 1 − P( X  5)
 e −12 12 0 e −12 121 e −12 12 4 
= 1 −  + + ... +  = 0.0959
 0! 1! 4! 
Next: Use Cumulative Poisson Probabilities Table for Example 8
 Using The Cumulative Poisson Probabilities Table
29

STA466 Mathematical Statistics I [email protected]

30
Example 5 Solution (using Table 2)
X ~ Poi (2.4)

(a) P(X=0) = P(X ≥ 0) - P(X ≥ 1)

= 1.0000 – 0.9093
= 0.0907

(b) P(X ≥ 4) = 0.2213

(c) P(X ≥ 5) = 0.0959

Note: For Poisson distribution,

it is easier to use Table 2
31
Example 6
Cars arrive at a petrol station at an average rate of 30 per hour.
Assuming that the cars arriving at the petrol station occur at
random, find the probability that:
a) No car arrives during a particular 5 minutes interval.

b) More than 3 cars arrive in 10 minutes interval.

c) Less than 5 cars during a period of half an hour.

 Example 6 Solution
32

Let X= Number of cars arrive at a petrol station (per

hour)
λ=30 (per hour), X~Poi(30)

30  5
(a) new = = 2.5 (for 5 minutes)
60

X~Poi(2.5)
e −2.5 2.5 0
P( X = 0) = = 0.0821
0!
Example 6 Solution cont.
33

(b)  new = 2.5  2 = 5 (for 10 minutes)

X~Poi(5)
0.7350
P( X  3) = P( X  4) = 0.9897
(c)
new = 30  2 = 15 (for 30 minutes)
X~Poi(15)

P( X  5) = 1 − P( X  5)
= 1 − 0.9991 = 0.0009
 THE BERNOULLI AND
34
BINOMIAL DISTRIBUTIONS

[email protected]
 THE BERNOULLI AND
35
BINOMIAL DISTRIBUTIONS

[email protected]
 POISSON APPROXIMATION
36
TO BINOMIAL DISTRIBUTION

[email protected]
37
Using the Poisson Distribution as an
Approximation to the Binomial Distribution
When n is large (n > 50) and p is small (p < 0.1),
the binomial distribution X~Bin(n, p) can be
approximated using Poisson distribution with the same
mean, i.e. X~Poi(np).

The approximation gets better as n gets larger and p gets

smaller.
38 Example 7
Eggs are packed into boxes of 500. On average, 0.7% of the
eggs are found to be broken when the eggs are unpacked.
Find the probability that in a box of 500 eggs,
(a) exactly three are broken,
(b) at least two are broken.
 Example 7 Solution
39
X = Number of broken eggs in a box of 500
p = P(egg is broken)= 0.007, n = 500
so X~Bin(500, 0.007)
Since n > 50 and p < 0.1, hence use a Poisson approximation
E(X) = np = (500)(0.007) = 3.5
X~Poi(3.5).
(a) e −3.5 3.53
P( X = 3) = = 0.2158
3!

(b) P( X  2) = 1 − (P( X = 0) + P( X = 1) )
 e −3.5 3.5 0 e −3.5 3.51 
= 1 −  +  = 0.8641
 0! 1! 
or you may use the table straight away and get the answer
40 SUMMARY

STA466 Mathematical Statistics I [email protected]