You are on page 1of 23

Poisson Distribution Function

(i) 𝒏, the number of trials is indefinitely large, i.e., 𝐧 → ∞.


(ii) 𝒑, the constant probability of success for each trial is indefinitely small,
i.e., 𝒑 → 𝟎.
(iii) 𝒏𝒑 = 𝝀 is finite, where 𝝀 − is a positive real number.
𝜆 𝜆
Thus, 𝑝 = 𝑞 =1−
𝑛 𝑛

The probability of 𝒙 successes in a series of 𝒏 independent trials is


𝑛
𝑏 𝑥; 𝑛, 𝑝 = 𝑥
𝑝 𝑥 𝑞 𝑛−𝑥 ; 𝑥 = 0, 1,2 … . . 𝑛
Poisson Distribution as a limiting form the B.D.
𝑛
𝑏 𝑥; 𝑛, 𝑝 = 𝑥
𝑝 𝑥 𝑞 𝑛−𝑥
𝑛
= 𝑥
𝑝 𝑥 (1 − 𝑝)𝑛−𝑥 (∵ 𝑞 = 1 − 𝑝)
𝑛
= 𝑥
𝑝 𝑥 (1 − 𝑝)𝑛 (1 − 𝑝)−𝑥
𝑛 𝑝 𝑥
= 𝑘
(1 − 𝑝)𝑛
1−𝑝
let 𝑛𝑝 = 𝜆 ⇒ 𝑝 = 𝜆/𝑛
𝜆 𝑥
𝑛 𝑛−1 𝑛−2 ….(𝑛−𝑥+1) 𝑛 𝜆 𝑛
= 𝜆 𝑥
1−
𝑥! 1− 𝑛
𝑛

1 2 𝑥−1
𝑛⋅ 𝑛 1−𝑛 ⋅ 𝑛 1−𝑛 …. 𝑛 1− 𝑛 𝜆𝑥 𝜆 𝑛
= 𝜆 𝑥
1 −
𝑥! 1−𝑛 𝑛𝑥 𝑛

Limiting form of the B.D. under the above conditions:


1 𝜆 𝑛
As 𝑛→∞ ⇒ → 0 and 1− → 𝑒 −𝜆
𝑛 𝑛
𝑒 −𝜆 𝜆𝑥
lim 𝑏 𝑥; 𝑛, 𝑝 = ; 𝑥 = 0, 1,2 … . . ∞
𝑛→∞ 𝑥!
𝑝→0
Definition: A r.v. X is said to follow Poisson distribution if it assumes
only non-negative values and its p.m.f. is given by:

𝑒 −𝜆 𝜆𝑥
; 𝑓𝑜𝑟 𝑥 = 0, 1,2 … . . ∞; 𝜆>0
𝑥!
𝑃 𝑋=𝑥 =
0 ; 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒

𝜆 - is the parameter of the distribution.


1. Poisson distribution occurs when there are events which do not
occur as outcomes of a definite number of trials of an experiment
but which occur at random points of time and space where in our
interest lies only in the number of occurrences of the event, not in
the non-occurrences.
2. Some instances:
(i) Number of deaths from a disease.
(ii) Number of defective materials in a packing manufactured by a good
concern.
(iii) Number of air accidents in a unit of time.
(iv) Number of printing mistakes at each page of the book.
(v) The emission of radioactive particles.
Moments of Poisson D.F.
Non- Central moments: 𝜇1′ = 𝐸 𝑋 = 𝜆
𝜇2′ = 𝐸 𝑋 2 = 𝜆2 + 𝜆
𝜇3′ = 𝐸 𝑋 3 = 𝜆3 + 3𝜆2 + 𝜆
𝜇4′ = 𝐸 𝑋 4 = 𝜆4 + 6𝜆3 + 7𝜆2 + 𝜆
Central moments:
′ ′ 2
𝜇2 = 𝑉 𝑥 = 𝜇2 − (𝜇1 ) = 𝜆
′ ′ ′ ′ 3
𝜇3 = 𝜇3 − 3𝜇1 𝜇2 + 2(𝜇1 ) = 𝜆
′ ′ ′ ′ ′ 2 ′ 4
𝜇4 = 𝜇4 − 4𝜇1 𝜇3 + 6𝜇2 (𝜇1 ) − 3(𝜇1 ) = 3𝜆2 + 𝜆

𝜇32 1 1
𝛽1 = = and 𝛾1 = 𝛽1 =
𝜇23 𝜆 𝜆
𝜇4 1 1
𝛽2 = =3+ and 𝛾2 = 𝛽2 − 3 =
𝜇22 𝜆 𝜆
M.G.F. of Poisson Distribution Function

𝑀𝑋 𝑡 = 𝐸 𝑒 𝑡𝑋 = 𝑒 𝑡𝑥 𝑃(𝑋 = 𝑥)
𝑥=0

𝑒 −𝜆 𝜆𝑥
∞ 𝑡𝑥
= 𝑥=0 𝑒
𝑥!

𝜆(𝑒 𝑡 −1)
= 𝑒
Example
For the case of the thin copper wire, suppose that the number of flaws follows
a Poisson distribution with a mean of 2.3 flaws per millimeter. Determine the
probability of exactly 2 flaws in 1 millimeter of wire.
Also find the probability of 10 flaws in 5 millimeters of wire.
Solution
Let X denote the number of flaws in 1 millimeter of wire.
Then, 𝐸 𝑋 = 𝜆 = 2.3 and probability of exactly 2 flaws in 1 millimeter of wire is
𝑒 −2.3 2.32
𝑃 𝑋=2 = = 0.265
2!

Determine the probability of 10 flaws in 5 millimeters of wire.


Let 𝑌 denote the number of flaws in 5 millimeters of wire. Then, Y has a Poisson distribution
with
𝐸 𝑋 = 𝜆 = 5𝑚𝑚 × 2.3𝑓𝑙𝑎𝑤𝑠/𝑚𝑚 = 11.5 𝑓𝑙𝑎𝑤𝑠

𝑒 −11.5 11.510
∴ 𝑃 𝑌 = 10 = = 0.113
10 !
Example
If the average number of claims handled daily by an insurance company is 5,
what proportion of days have less than 3 claims? What is the probability that
there will be 4 claims in exactly 3 of the next 5 days? Assume that the number
of claims on different days is independent.
Solution:
Because the company probably insures a large number of clients, each having a small
probability of making a claim on any given day, it is reasonable to suppose that the
number of claims handled daily, call it 𝑋, is a Poisson random variable.
Since 𝐸(𝑋 ) = 5,
the probability that there will be fewer than 3 claims on any given day is

𝑃{𝑋 < 3} = 𝑃{𝑋 = 0} + 𝑃{𝑋 = 1} + 𝑃{𝑋 = 2}


51 2
= 𝑒 −5 +𝑒 −5 +𝑒 −5
1! 2!
𝑃{𝑋 < 3} ≈ 0.1247
Since any given day will have fewer than 3 claims with probability .125, it follows, from
the law of large numbers, that over the long run 12.5 percent of days will have fewer than
3 claims.
It follows from the assumed independence of the number of claims over
successive days that the number of days in a 5-day span that has exactly 4
claims is a binomial random variable 𝒀, with parameters 𝒏 = 𝟓 and
𝒑 = 𝑷{𝑿 = 𝟒}.
54
𝑃 𝑋=4 = 𝑒 −5 ≈ 0.1755
4!

it follows that the probability that 3 of the next 5 days will have 4 claims is,
𝑌~𝐵(5, 0.1755)

5 3 2
P Y=3 = 3
0.1755 0.8245 = 0.0367
Example
It is known that the probability of an item produced by a certain machine will
be defective is 0.05. If the produced items are sent to the market in a packet
of 20, find the number of packets containing at least, exactly and at most 2
defective items in a consignment of 1000 packets using (i) Binomial
distribution and (ii) Poisson approximation.
Example
Find the probability of 5 or more telephone calls arriving in a 9 min period in
switch-board, if the telephone calls that are received at the rate of 2 calls in
every 3 minute period.
Recurrence relation for Poisson Distribuiton
𝒆−𝝀 𝝀𝒙
𝑷 𝑿=𝒙 = ; 𝒙 = 𝟎, 𝟏, 𝟐, … … … , ∞
𝒙!

𝒆−𝝀 𝝀𝒙+𝟏
𝑷 𝑿=𝒙+𝟏 = ; 𝒙 = 𝟎, 𝟏, 𝟐, … … … , ∞
𝒙+𝟏 !

p 𝑥+1 𝜆
∴ =
p(𝑥) 𝑥+1
p 𝑥+1 = 𝜆 p(𝑥)
𝑥+1
p 0 = 𝑒 −𝜆 or 𝑥=𝜆
The remaining probabilities can be obtained as:

𝜆
p 1 = p 𝑥+1 𝑥=0 = p(0)
𝑥+1 𝑥=0

𝜆
p 2 = p 𝑥+1 𝑥=1 = p(1)
𝑥+1 𝑥=1

𝜆
p 3 = p 𝑥+1 𝑥=2 = p(2)
𝑥+1 𝑥=2

and so on.
Expected frequencies:

𝑓 0 = 𝑁 × 𝑃(𝑋 = 0)

𝑓 1 = 𝑁 × 𝑃(𝑋 = 1)

𝑓 2 = 𝑁 × 𝑃(𝑋 = 2)

and so on..
Example
After correcting 50 pages of the proof of a book, the proof reader finds
that there are on the average, 2 errors per 5 pages. How many pages would one
expect to find with 0, 1, 2, 3 and 4 errors in 1000 pages of the first print of the
book ?
Let, 𝑿 – denote the number of errors per page.

Average number of errors per page is given by:


2
𝜆 = = 0.4
5

𝒆−𝝀 𝝀𝒙
∴ 𝑋~𝑃 0.4 ⇒ 𝑃 𝑋 = 𝑥 = ; 𝒙 = 𝟎, 𝟏, 𝟐, … … …
𝒙!

Expected number of pages with 𝑿 − errors per page in a book of 1000 pages are
:
𝑒 −0.4 0.4 𝑥
1000 × 𝑃 𝑋 = 𝑥 = 1000 × ; 𝑥 = 0, 1, 2, … …
𝑥!
Using the recurrence relation of P.D.,

No. of errors Probability Expected number of pages


per page (𝑿) 𝑷(𝑿 = 𝒙) 𝟏𝟎𝟎𝟎 × 𝒑(𝒙)

0 𝑝 0 = 𝑒 −0.4 = 0.6703 670.3 ≃ 670


0.4
1 p 1 = p 0 = 0.26812 268.12 ≃ 268
0+1
0.4
2 p 2 = p 1 = 0.05362 53.624 ≃ 54
1+1
0.4
3 p 3 = p 2 = 0.007129 7.1298 ≃ 7
2+1
0.4
4 p 4 = p 3 = 0.00071298 0.71298 ≃ 1
3+1
Example Problems
1. Past Experience reveals that Mr. Rana can hit the target 3 times out of 5 shots.
Find the probability of hitting the target,
(i) 4 out of 6 shots. [Ans:0.31104 ]
(ii) at least 2 times out of six shots. [Ans:0.959]
(iii) not more than once out of six shots. [Ans:0. 0409]

2. A box contains 20 capacitors, 4 of which are defective. Five capacitors are


selected randomly. What is the chance that,
(i) 3 out of 5 selected capacitors are defective? [Ans:0.0512 ]
(ii) There is no defective capacitors among these 5? [Ans:0.3276 ]
(iii) There is at least one defective? [Ans:0.6723 ]
3. The fist print of a new book consisting of 750 pages has on an average 1 error in five
pages. Find the number of pages which have 0, 1, 2, 3 errors per page in the whole book.

Ans: X 0 1 2 3
No. of Pages 614 123 12 1

4. A razor blade manufacturing company has a chance of 1 defective blade in 500 blades.
The blades are marketed in packets of 5 blades. One hundred packets are supplied to
retailer. Find the number of packets which are likely to have no, 1 or 2 defective blades.
Ans:
X 0 1 2
No. of
99 1 0
Packets
5. Each sample of water has a 10% chance of containing a particular organic
pollutant. Assume that the samples are independent with regard to the
presence of the pollutant. Find the probability that in the next 18 samples,
(i) exactly 2 contain the pollutant, (ii) atleast four samples contain the
pollutant. [Ans: 0.284, 0.098]

You might also like