Poisson Distribution

A discrete random variable, X, is said to have a Poisson distribution if its probability function
e −  x
is given by P ( x ) = ; for x = 0,1, 2, ...,  where, e=2.718 and  is the parameter which is
x!
the average number of success.

Poisson distribution is a limiting case of binomial distribution under the following conditions:

(i) The number of trials, n, of a random experiment is independently large, i.e. n → 

(ii) The probability of success, p, is constant for each trial and it is indefinitely small, i.e.
p → 0.

(iii) The mean, np =  , is finite.

Examples.

1. Number of defective materials produced in an industry.

2. Number of printing mistakes at each page of a book.

3. Number of telephone calls received at a particular time in a telephone exchange.

4. Number of air accidents in some unit of time.

5. Number of accidents at a crossing per hour during the busy time of a day.

Derivation. Poisson distribution can be derived from binomial distribution under the
following conditions:

(i) The number of trials, n, is very large, i.e. n → 

(ii) The probability of success, p, is very small, i.e. p → 0 .

(iii) np =  , is finite constant, that is average number of success is finite.


We have np =   p =  n and  q = 1 − = 1− p
n

The probability function of binomial variate X with parameters n and p, is given by

x n− x
n!   
P ( X = x ) = Cx p q
n x n− x
=   1 − 
x !( n − x ) !  n   n
−x n
x    n! 
= 1 −  1 −  x
x !  n   n  n ( n − x )!

Page 1 of 3
 −x
 
For fixed x, lim 1 −  =1
n →
 n
n! n ( n − 1)  n − ( x − 1)  ( n − x ) !
and lim = lim
n → n x ( n − x ) ! n → n x ( n − x )!
 1  2   x −1 
n x 1 −  1 −  1 − 
n  n  n 
= lim  
n → nx
=1

n −
  n
while, lim 1 −  = e n = e− 
n →
 n

e−   x
Hence, lim P ( x ) = ; x = 0,1, 2, ,
n → x!

which is the probability function of Poisson distribution with parameter  .

Mean
 
e −  x 
xx 
x−1
E (x ) =  x  P(x ) =  x  = e −  = e −  
x =0 x =0 x! x =0 x ( x − 1)! x =1 ( x − 1)!

  3 2

= e −  1 +  + + + ... = e −   e  = 
 2! 3! 

Variance

( )
V (x) = E x 2 − E(x) = Ex(x − 1) + x− E(x) = Ex(x − 1)+ E(x) − E(x)
2 2 2

 
e −  x 
x( x − 1)x 
x −2
Ex( x − 1) =  x( x − 1)  P( x ) =  x( x − 1)  = e −  = e − 2 
x =0 x =0 x! x =0 x ( x − 1)( x − 2 )! x = 2 ( x − 2 )!

    2 3
= e − 2 1 +  + + + ... = e − 2  e  = 2
 2! 3! 

V (x ) = 2 +  − 2 =  .

Therefore, for Poisson distribution, Mean=Variance=  .

Page 2 of 3
Ex. 1. The electric bulbs produced in an industry are found defective at the rate of 2 percent.
In an hour of a day the industry produces 500 bulbs. Find the probability that, in that hour (i)
4 defective bulbs are produced, (ii) at least 2 defective bulbs are produced, (iii) at best 2
defective bulbs are produced.

Sol. Let X be the number of defective bulbs. The average number of defective bulbs produced
in an hour is np =  . Here, n=500 and p=0.02. Hence,  = 500  0.02 = 10 .

The probability function of X is

e −  x e −1010 x
P( X = x ) = = ; x = 0,1, 2, ...
x! x!

e −1010 4
(i) P ( X = 4 ) = = 0.01892
4!

 e −1010 0 e −10101 
(ii) P( X  2) = 1 − P( X  2) = 1 − P( X = 0) + P( X = 1) = 1 −  +
 0! 1! 

= 1 − 0.000499 = 0.9995

e −1010 0 e −10101 e −1010 2

(iii) P( X  2 ) = P( X = 0 ) + P( X = 1) + P( X = 2) = + + = 0.00277
0! 1! 2!

Ex. 2. The average printing mistakes per page of a book is 1. Find the probability that, in the
book there will be (i) 10 mistakes per page, (ii) at best 10 mistakes per page, (iii) at least 10
mistakes per page.

Sol. Let X be the number of mistakes in a page. Here,  = 1 . The probability function of X is

e −  x e −11x
P( X = x ) = = ; x = 0,1, 2, ...
x! x!

e −1110
(i) P( X = 10 ) = = 0.000000101
10!

(ii) P( X  10 ) = P( X = 0 ) + P( X = 1) + P( X = 2 ) + ... + P ( X = 10 )

e −110 e −111 e −112 e −1110 10 11 12 110 

= + + + ... + = e −1  + + + ... +  = e −1  2.71828 = 0.9999993
0! 1! 2! 10!  0! 1! 2! 10!

9
(iii) P( X  10 ) = 1 − P( X  10 ) = 1 −  P( X ) = 1 − 0.9999993 = 0.0000003
X =0

Page 3 of 3