Poisson The distribution
The Poisson distribution is, like the Binomial distribution, a theoretical discrete
distribution that is used to calculate the probability that a random event occurs a given number of
times. Given a Poisson variable X, the values assumed by X have probabilities that can be
generated by its probability mass function
P[ X = x] =
e x
for x = 0, 1, 2, 3,
x!
where X is defined as the number of occurrences of the variable and is the mean
number of occurrences over a specific interval (explained below). In fact, is the only
parameter of the Poisson distribution and also happens to be the variance of the distribution. We
hence write
X ~ Po( )
Unlike the case for the binomial distribution, we do not have an upper limit for the
number of occurrences of a randomly occurring event (which, as you will see
further, makes a lot of sense). In some books, the parameter is used instead of
.
The conditions for applying the Poisson distribution to a given problem are
1.
The occurrences of the event are completely independent of one another (that is,
they are completely random)
2.
The number of occurrences in a given space-time interval is independent of those
occurring in another interval.
An example of a randomly occurring event is the number of earthquakes in Turkey in a one
year-period. It is seen in this example that the time interval has to be defined even though we
know that our variable is the number of earthquakes in Turkey. This is what a space-time
interval is all about! Other variables could be
1.
2.
3.
4.
Number of breaks in a white line on a 1-kilometre road (length)
Number of misprints per page in a book (area)
Number of micro-organisms in 10 cm3 of liquid (volume)
Number of customers arriving in a bank per minute (time)
Now that we understand the concept a of space-time interval, let us have a look at an elementary
example.
Example 1
The number of emergency admissions each day to a hospital is found to have a Poisson
distribution with mean 2.
(a)
(b)
Evaluate the probability that, on a particular day, there will be no emergency admissions.
At the beginning of one day the hospital has 5 beds available for emergencies. Calculate
the probability that this will be an insufficient number for the day.
Calculate the probability that there will be exactly 3 admissions altogether on two
consecutive days.
(c)
Solution
We should always start by defining our variable since its distribution varies according to its
space-time interval. So if X = number of emergency admissions per day, we write X ~ Po(2).
e 2 2 0
= 0.135
0!
(a)
P[ X = 0] =
(b)
If the number of beds is insufficient, we can easily deduce that the number of
emergencies exceeded 5. Hence we have to calculate P[ X > 5] . Unfortunately, that
would mean calculating an infinite number of probabilities since the Poisson variable
has no upper limit for its number of occurrences. We revert to the theory of
probability which states that the sum of all probabilities in a given experiment must
be 1. It follows that
P[ X > 5] = 1 P[ X 5] .
Now, P[ X 5] = P[ X = 0] + P[ X = 1] + ....+ P[ X = 5]
e 2 2 0 e 2 21 e 2 2 2 e 2 2 3 e 2 2 4 e 2 2 5
+
+
+
+
+
0!
1!
2!
3!
4!
5!
4 2 4
= e 2 1 + 2 + 2 + + + = 0.983
3 3 15
=
Thus, P[ X > 5] = 1 0.983 = 0.017 .
(c)
This part of the question shows how flexible the Poisson distribution can be! The
space-time time interval controls the parameter of the distribution. We have to define
a new Poisson variable (lets call it Y) defined as
Y = number of emergency admissions for a two-day period
so that Y ~ Po(4).
P[Y = 3] =
e 4 4 3
= 0.195
3!