0% found this document useful (0 votes)
16 views47 pages

Poisson Distribusion Latest

The document discusses the Poisson distribution and its properties and applications. It provides examples of modelling real-world events using the Poisson distribution such as births in a maternity ward and cyclists passing a post office. It also discusses using the normal distribution to approximate the Poisson distribution when the parameter is large and compares the binomial and Poisson distributions.
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
16 views47 pages

Poisson Distribusion Latest

The document discusses the Poisson distribution and its properties and applications. It provides examples of modelling real-world events using the Poisson distribution such as births in a maternity ward and cyclists passing a post office. It also discusses using the normal distribution to approximate the Poisson distribution when the parameter is large and compares the binomial and Poisson distributions.
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 47

Poisson Distribution

Conditions of Poisson distribution

1.Evens are discrete (You can count them)

2.Events cannot happen at the same time

3.Events are indepenedent


Poisson Parameter
The rate
Number of events per time period
Lambda λ
Mean = λ
Variance = λ
Standard deviation
Requirements and Assumptions of Poisson
• Individual events cannot happen at the same time

• Individual events are independent

• The probability that an event occurs does not


depend on how long it has been since the last event
Modelling with the Poison distribution
Describes the
1. The Number of occurrences in a fixed period
of time or space
2. Events occur independently of one another
3. at random
4. at a constant average
Examples of Poisson Districbution

1.Radioactive emissions
2.Traffic passing a fixed point
3.Telephone calls or letters arriving
4.Accidents occurring
Examples of Poisson Distribution
1.Maternity ward – How many births at a night
2.Total deliveries per year – 1000
3.Average births per night – 1000/365
4.Probabilities of number of babies delivering
at night
5. P(X=0) , P(X=1), P(X=2) etc
Examples of Poisson Distribution
= = 0.064
Similarly,
0.177
0.242
0.221
Examples of Poisson Distribution

Similarly,

] * 365
= 1- [0.856] * 365
= 0.144 * 365
= 52.56
Greatest number of deliveries once an year at least once
X Mean P
0 2.74 0.06457 Minimum deliveries
1 2.74 0.176923 = 1/365
2 2.74 0.242384
3 2.74 0.221378 = 0.00273
4 2.74 0.151644
5 2.74 0.083101
6 2.74 0.037949
7 2.74 0.014854
8 2.74 0.005088
9 2.74 0.001549
10 2.74 0.000424
Why pattern does not follow the poisson Distribusion
If the deliveries are not random through out the
day

E.g. Specially the sesarians are done in the day


time
Condition of the poison distribution
1. Events occur at random
2. Events occur independently of one
another
3. The average rate of occurrences
remains constant
4. There is zero probability of
simultaneous occurrences
Example Poison Distribution
1. Cyclists passing a remote village post-office during the day,
Average 2/ hour
a) Probability that
i). No cyclist passes
P(X=0) = e(-2) = 0.1353

ii). P(X>3) = 1- P(X<=3) = 1 – (0.1353+0.270+0.270+0.180)


= 1- 0.8553 = 0.1447
Example Poison Distribution
b) Probability that exactly one passes by 20 min
New lambda = 2/3 = 0.66

x = 0.342 (3 s.f.)
Example Poison Distribution
c) What is the probability that more than three cyclists pass by in an
hour exactly once in a 6-period?

as the there are 6 trials,

Y = number of times that more than 3 cyclists pass by an hour exactly


once in a 6-hour period

Y ~ B(6, 0.1429)

P(Y=1) = nCr Pr Q(n-r) = 6C1 P1 Q(6-1) = 6 * 0.14291 (1-0.1429)5 = 0.397 (3 s.f.)


Example Poison Distribution
2). The number of boats arriving in a 15 min period with average
a) . Find the probability that exactly 6 boats will arrive of an hour
P(X=6) = = 0.161

b) If twice arrive in second hour

P(2 boats arrive I half hour/ then 4 boats arrive in next half hour) = x = 0.0376

Conditional probability = P(2 boats arrive I half hour/ then 4 boats arrive in next half hour)

(6 boats arrive in an hour)


= 0.0376 / 0.161
= 0.234
•Poison provides a good approximation
• When n is large under certain conditions
Approximati • If X ~ B(n,p) with n large (n>50)
on Involving •
Poisson • P close to 0 -------- (np<5)
Distribution
• Then X ~ approximately Po(λ) with λ =
np

•Pls see the excel sheet –--------------


Approximation Involving Poisson
Distribution
•Example 1
•The probability that a component coming off a
production line is faulty is 0.01
•a) If a sample size 5 is taken, find the probability
that exactly one of the component is faulty,
Approximatio
n Involving •P(X=1) = 5 * 0.01 * 0.994 = 0.0475

Poisson
Distribution •b) What is the probability that a batch of 250 of these components has more than
•3 faulty components in it?

•If Y – number of faulty components in the batch X ~ B(250, 0.01) ~ Approx Po(2.5)

•So, the Poisson approximation to find the probability P(Y>3) = 1-[P(Y<=3) ]


•Use for lambda >15

•P(X≥15) , P(6<X<14)

Normal
•Normal distribution uses standard
Approximati deviation to calculate the z-score
on to Poisson
Distribution
•** You must also include the continuity
correction
•Example
•If X ~ Po(16) calculate
• Using Poison Distribution
Normal • = =
Approximati
on to Poisson •b) Using Normal Distribution
Distribution • for the Poisson
•Use N(16,16)

•Continuity correction factor = ~ Approxi

•Y is the approximating normal,


•Example
•Continuity correction factor = ~ Approxi
•Y is the approximating normal,
Normal
Approximati
on to Poisson • ]
Distribution • = P[]

• = P(-1.375 < Z <-0.125)
• =
• = 0.9155 – 0.5498 = 0.366
Normal Approximation to Poisson Example

Distribution
Normal Approximation to Poisson Example

Distribution
April 12, 2024
Normal distribution as an approximation for
the Poisson distribution
You may use the normal distribution as an approximation for
the Poisson distribution, provided that its parameter (mean) λ
is sufficiently large for the distribution to be reasonably
symmetrical and not positively skewed.

As a working rule λ should be at least 15. If λ = 15,


mean = 15 and standard deviation = 15 = 3.87 (to 3 s.f.).
Normal distribution as an approximation for
the Poisson distribution
A normal distribution is almost entirely contained within 3
standard deviations of its mean and in this case the value 0 is
between 3 and 4 standard deviations away from the mean
value of 15.

The parameters for the normal distribution are then


Mean: µ = λ Variance: σ 2 = λ
Poison Approximation to the Binomial Distribution

The Poisson distribution may be used as an


approximation to the binomial distribution,
B(n, p),
• when: » n is large (typically n > 50)
• p is small (and so the event is rare)
• np is not too large (typically np < 5).
Approximation of Distributions
The sum of two or more poison distributions

Assuming that the flow of vehicles, into and out of town,


can be modelled by independent Poisson distributions,
you can model the flow of vehicles in both directions as
follows
• Let X represent the number of vehicles travelling into
town at 3 pm, then X ∼ Po(3.5).

• Let Y represent the number of vehicles travelling out of


town at 3 pm, then Y ∼ Po(5.7).

• Let T represent the number of vehicles travelling in


either direction at 3 pm, then T = X + Y.
• You can find the probability distribution for T as follows.
P(T = 0) = P(X = 0) × P(Y = 0)
= 0.0302 × 0.0033
= 0.0001

P(T = 1) = P(X = 0) × P(Y = 1) + P(X = 1) × P(Y = 0)


= 0.0302 × 0.0191 + 0.1057 × 0.0033 = 0.0009

P(T = 2) = P(X = 0) × P(Y = 2) + P(X = 1) × P(Y = 1) + P(X = 2)


× P(Y = 0)

= 0.0302 × 0.0544 + 0.1057 × 0.0191 + 0.1850 × 0.0033


= 0.0043
Since this is very much time consuming,
We will model the distribution when two independent Poisson variables of
λ and μ.
Then,
Po(λ) & Po(μ) ------------ T = X + Y , T  Po(λ + μ)
So,
x =

x = x 9.2
P(T>10) = 1- P(T<=10)
= 1 – 0.6820
= 0.318
Question 02.
Average Death causes per year due to a rare decease by
country,

Sweden ------- 2.0


Norway ------- 0.8
Finland ------- 0.5

 Deaths causes random


 Cases are independent of one another
Question 02.
Average Death causes per year due to a rare decease by country,
Since, three distributions are independent we can add the
average together to form a new random variable T

Po(2 ), Po(0.8), Po(0.5)

T = 2+0.8+0.5 = 3.3, there new random variable

Po(3.3)
Question 02.
Po(3.3)

P(T>=4) = 1- P(T<4 )

= 1 – 0.5803

= 0.4197
Question 03.

• On a lonely Highland road in Scotland,


• cars are observed passing at the rate of 6 per
day and lorries at the rate of 3 per day.

• On the road is an old cattle grid that will soon


need repair.
Question 03.

The local works department decides that if the probability


of more than 2 vehicles per hour passing is less than 1%
then the repairs to the cattle grid can wait until next spring,
otherwise it will have to be repaired before the winter.
When will the cattle grid have to be repaired?
Question 03.
Hypothesis test for the mean of a Poisson
distribution
Hypothesis test for the mean of a Poisson
distribution
Hypothesis test for the mean of a Poisson
distribution
Hypothesis test for the mean of a Poisson
distribution
Hypothesis test for the mean of a Poisson
distribution
H0 : U = 5.6
H1: U <5.6
Hypothesis test for the mean of a Poisson
distribution

You might also like