Binomial, Poisson & Normal Distribution

Binomial, Poisson & Normal Distribution
(For M.B.A. I Semester)
by Prof. Nitin Karoulia

9033099006
•
Combination (Definition). A combination of n different
objects taken r at a time, denoted by or n ,is a
selection of only r objects out of the
r
n objects, without any regards to the order of arrangement.
Theorem 12.6. The number of different combinations of n

different objects taken r at a time, without repetition, is
= n = n! ;r<n
r r ! (n – r) !
Binomial Distribution
Binomial distribution was discovered by Swiss Mathematician James Bernoulli, so this
distribution is called as ‘Bernoulli Distribution’ also. This is a discrete frequency
distribution.
Assumptions of a Binomial Distribution
1. The random experiment is performed repeatedly with a fixed and finite number of
trials. Number of trials is denoted by ‘n’.
2. There are two mutually exclusive possible outcomes on each trial, which are known
as ‘success’ and ‘failure’. Success is denoted by p whereas failure is denoted by q, and
p+ q = 1, or q = 1 - p.
3. The outcomes of any given trial does not affect the outcomes on subsequent trials,
means the trials are independent.
4. The probability of success and failure (p & q) remains constant from trial to trial. If in
any distribution the p & q does not remain constant, that distribution cannot be a
binomial distribution. For example, the probability of getting head (or tail) must
remain the same in each toss, i.e.1/2. Similarly the probability of drawing 4 balls
from a bag containing 6 red and 10 white balls does not change in successive draws
with replacement, hence it will be called as binomial distribution. But, in contrary to
this, if balls are not replaced after each trial, then it will not be a binomial
distribution.
• number of success (r) in descending order (…, 4, 3, 2, 1, 0) can be
Binomial distribution is the expansion of , from which the probability of
determined. For determining the probability in ascending order we use

the expansion of . For example if n = 5, p = probability of success
(Heads), q = probability of failure (trials) then the expansion of & may
be given as follows :
Expected Frequencies of Binomial Distribution
For determining the expected frequencies, the probability of each item is
multiplied by total number of trials (N)
N(p + q)n or N(q + p)n
N = Total number of trials.
n = Number of independent events (coins, dices, etc.)
Constants of Binomial Distribution
Binomial Distribution is a special type of distribution, the constant values of it
i.e. Mean (X) and Standard Deviation (s) Can be computed by using the
following formulae_ :
X = np
σ = √npq
Illustration 1 : Six coins are thrown simultaneously. Find the chance of
obtaining (i) no head, (ii) at least one head, (iii) exactly two heads, (iv) not
more than two heads; (v) more than 3 heads.
Solution :
Poisson Distribution
•
The Poisson Distribution was discovered by French mathematician
Simon Denis Poisson. It is a discrete probability distribution.
Meaning :
In binomial distribution if the value of n is very large (n = ) and the value of p
is too small (p 0) and np is finite number, in this situation the Binomial
distribution is not suitable to be used. In other words, the Poisson
distribution is applicable where the successful events in the total events
are few.
Situations where Poisson Distribution is applicable :
1) Number of defective blades out of total blades produced in a factory.
2) Number of goals scored at a football match, wherein number of attempts
(n) may be a lot of but the success (p) are few.
3) Number of mistakes found in the pages of a book published by a repute
press.
4) Number of telephone calls done during every 5 minutes by a
businessman.
5) Number of typing errors per page in a typed material.
6) No. of accidents met by a taxi driver in a year.
7) The arrival of customers arriving per hour at the super-market.
8) The arrival of trains arriving per 10 minutes in the Railway Yard.
9) Number of flying bombs dropped per year in a small area.
Poisson Probability Functions
The probability of ‘r’ success in poison distribution is determined by the
following formula :
Here P(r) = probability of ‘r’ time success

e = It is mathematical constant value which is derived from the natural law of
exponential growth, its value is 2.71828.
• m = np.
m = Arithmetic mean, if ‘m’ is not given, it may be computed by using formula
Poisson Distribution as a Limiting Form of Binomial Distribution Under the

following situation we use Poisson distribution in place of Binomial
distribution :
i. When number of trials (n) are unlimited, means n →
ii. When the probability of success, ‘p’→ 0; and
iii. When np = m is finite.
Constants of Poisson Distribution :
iv. Arithmetic mean = np = m
v. Standard Deviation (σ) = √np = √n
Here variance = m
Hence, in Poisson distribution the value of mean and variance are equal.
Expected Frequency
The expected frequency in Poisson distribution can be determined by
multiplying ‘N’ with all the values,
Illustration 1:
In a certain Poisson frequency distribution the frequency corresponding to 2
successes is half the frequency corresponding to 3 successes. Find its
mean and standard deviation.
Solution:
Illustration 2:
If a random variable x follows Poisson distribution such that P(x = 1) = P(x
= 2), find the mean and variance of the distribution. Also find P(x = 0) and
P (x > 2).
Solution
Illustration 3:
If X follows Poisson distribution such that P(X = 2)= 9P(X= 4)+ 90P(X = 6), find
the mean and variance of X.
Solution :
Illustration 4:
Poisson distribution such that P(x = 1) = P(x = 2), find the mean and variance
of the distribution.
Solution :
Illustration 5:
The average number of customers, who appear at a counter of a certain bank
per minute is two. Find the probability that during a given minute----
i. No customer appears.
ii. Three or more customers appear.
Given e (- 2) = 0.1353
Solution:
Normal Distribution
Meaning : According to Ya-Lun-Chou : “The perfectly smooth and
symmetrical curve, resulting from the expansion of the binomial (p+ q)n
when n approaches infinity is known as the Normal Curve. Thus the
normal curve may be considered as the limit towards which the binomial
distribution approaches as n increases to infinity. Alternatively we may
say that the normal curve represents a continuous and infinite binomial
distribution, or simply a normal distribution”.
Normal distribution is a continuous probability distribution. During 18th
century Karl Friedrich Gauss (a German Mathematician/Scientist)
contributed an important role in the development of normal distribution,
so in his honor this distribution is often referred to as the ‘Gaussian
Distribution’. This distribution is also called as ‘Normal Law of Error’
because Gauss derived its equation from the study of error in repeated
measurements of the same.
When ‘n’, the number of trials is very large or infinite (n→ө) and neither ‘p’
nor ‘q’ is very small or nearer to equal then binomial distribution tends to
be a normal distribution.
Normal Distribution
Properties of Normal Curve
1. Shape : It is perfectly symmetrical and bell-shaped.
2. Position of Mean, Mode and Median : Mean, mode and median
remains equal in normal distribution, they are found in the mid of the
distribution, and distribute the area of curve in equally two parts.
3. Asymptotic : As the distance of the curve from the mean increases, the
curve comes closer and closer to the axis but never touches it.
4. Unimodal : It has only one mode so it is unimodal.
5. Continuous Distribution : It is a distribution of continuous variables.
6. Equidistance of Quartiles : The difference between ‘third quartile &
median’ (Q3 - M) and ‘median and first quartile’ (M - Q1) are equal
(Q3 - M) = (M - Q1)
7. Quartile Deviation and Probable Error : Quartile deviation is equal to
probable error, which is about ⅔ of standard deviation (QD = 0.6745σ)
Normal Distribution
8. Mean Deviation : The mean deviation about mean is ⅘ σ or 0.7979σ
9. Points of Inflection : The points where the curveity of normal curve
changes its direction are termed as ‘Points of Inflection’. These are found
at the difference between mean and one standard deviation (X
± σ). If straight lines are drawn from the points of inflection to the
axis (base line) both side _then the distance from the mean to those
points will be equal to X ± 1σ , X ± 2 σ , X ± 3 σ. Normal curve nearer to
mean are concave and nearer to two tails are convex.
10. Standard Deviation Determines the Width of Curve : If the value of
standard deviation is less the width of curve will also be less, if the
same way if the value of standard deviation is greater the width of the
curve will also be greater.
11. Parameters : There are two parameters of normal distribution :
Arithmetic Mean (N) & Standard Deviation (s).
12. Constants : The following are the constants of normal distribution :
Normal Distribution
Mean X or µ ; Standard deviation = σ
Variance = σ2 = µ2
Third central moment = µ3 = 0
Fourth central moment = µ4 = 3σ4 = 3µ2 2
Moment coefficient of Kurtosis β2 = µ4 = 3σ4 = 3
µ22 σ4
So, it is mesokurtic curve.
Moment coefficient of skewness β1 = µ32 = 0
µ 23
So, it is a symmetrical curve.
13. Since, the area of any continuous probability distribution is ë1í. So the
total area of a normal curve is ë1í. Area is probability, and the area left
side to mean equals to the area right side to mean, so area of both side
is 0.5 and 0.5.
Normal Distribution
14. Ordinate Relationship : The highest height is positioned at mean in
normal curve. It is called as maximum ordinate. At the distance of one
standard deviation from mean (X ± 1σ) the height of ordinate axis is
0.60653 or 60.65% of maximum ordinate axis in both sides. In the same
way the height of X ± 2σ ordinate axis is 0.13534 or 13.534% of
maximum ordinate axis. The table given below will show the percentage
of height
Distance from Mean (In Ratio of height from Percentage of height from
Standard mean ordinate mean ordinate
Deviation Unit)
0.5 .88250 88.25%
1.0 .60653 6.65%
1.5 .32465 32.47%
2.0 .13534 13.53%
2.5 .04394 4.39%
3.0 .01111 1.11%
Normal Distribution
15. Area Relationship : The important characteristic of normal curve is ‘Area
Relationship’. The area lying between the normal curve and the
horizontal axis (base line) is said to be the area under the normal curve,
which includes all the frequencies of distribution. The standard deviation
distributes the area under the normal curve as given in the following
table : Area Under Normal Curve
Distance from Mean (in Ratio of Area from Mean ordinate

standard
deviation unit)
Ratio of left side Ratio of right side Ratio of both side
area area areas
0.5 .19146 .19146 .38292 = 38.29%
1.0 .34134 .34134 .68268 = 68.27%
1.5 .43319 .43319 .86638 = 86.64%
2.0 .47725 .47725 .95450 = 95.45%
2.5 .39379 .39379 .98758 = 98.76%
3.0 .49865 .49865 .99730 = 99.73%
Normal Distribution
Standard Normal Distribution: The distribution which has mean 0 (X = 0) and
standard deviation 1 (s = 1) is called as standard normal distribution. The
area under the curve can be computed with the help of table on the basis of
distance between mean and different values of standard deviation. For this
purpose normal distribution is standardized by ‘Z’ transformation.
Standardized normal variable is denoted by ‘Z’. Normal distribution is
converted into standardized normal distribution as follows :
Z = X – X or x (Here x = X – X)
σ σ
The method of determining the value of Z is called as Z-transformation.
Process of Determining Area : First of all, the value of Z is computed by using
the formula given just above, thereafter by the help of table given in the
Annexure area is determined in the following manner :
The values of Z from 0.0 to 3.0 are given in the first column of the table. On
the basis of given value of Z respective area value is seen from the table. For
example Z= 2.54, then value before 2.5 and below 0.4 is 0.4945, which is
ratio of area. If the value of Z is positive then it is in the right side of
maximum ordinate (or mean), whereas if the value of Z is negative then it is
in the left side of maximum ordinate (or mean).
While determining the area it should be kept in the mind that the table
value always shows the area from arithmetic mean (Z= 0) to X (computed
value of Z). Hence the area should be determined accordingly.
Normal Distribution
It is well defined that the total area under normal curve is 1, which is found
0.5 in the right side and 0.5 in the left side of the mean.
Thus, the probability of having area more than 2.54
= (Total area in the normal curve Probability of having value of Z less than
2.54)
= 1 - .9945
= .0055
Determining the Area for Negative Value of Z----If the value of Z is negative
then this area will be situated left side to the mean. If area from mean
ordinate to Z= ----2.54 is to be determined then due to symmetry of
curve the area will be between mean ordinate and Z= 2.54 will be
0.4945.
If area is to be determined from the beginning to Z = -----2.54 then this area
will be computed by deducting 0.4945 (table value of Z when Z= ----2.54)
from 0.5
Normal Distribution
Illustration 1:
Suppose that the life of a gas cylinder is normally distributed with a mean of
40 days and standard deviation of 5 days. If at a time 10,000 cylinders are
issued to customers, how many will need replacement after 35 days ?
Solution:
Normal Distribution
Illustration 2:
The marks of the students are normally distributed. 10% get more than 75
marks and 20% get less than 40 marks. Find the mean and standard
deviation of the distribution.
The relevant extract of Area Table (under the Normal Curve) is given below :
Z 0.84 1.28 2.0
Area 0.2995(.3) 0.3997(.4) 0.4772
Normal Distribution
Solution:
Normal Distribution
Illustration 3:
The incomes of a group of 10,000 persons were found to be normally
distributed with mean equal to Rs. 750 and standard deviation equal to
Rs. 50. What was lowest income among the richest 250 ?
Solution :
Given : N = 10,000,
X = 750
σ = 50
Normal Distribution
Illustration 4:
In an intelligence test administered on 1000 students the average score was
42 and standard deviation 24. Find (i) the number of students exceeding a
score of 50, (ii) the number of students with score lying between 30 and
54; (iii) the value of score exceeded by the top 100 students.
Solution :
Normal Distribution
Normal Distribution
Illustration 5:
In a normal distribution 31% of the items are under 45; and 8% are over 64.
Find the mean and standard deviation of the distribution.
Solution :
31% items are less than 45, means the area left to the ordinate at 45 = 0.31.
So, the area from mean ordinate to the ordinate at 45 = 0.50 -- 0.31 = 0.19
for which Z’s value = ---- 0.5 (as per table)
Normal Distribution
Thank You

Binomial, Poisson & Normal Distribution

Uploaded by

Copyright:

Available Formats

Binomial, Poisson & Normal Distribution

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Binomial, Poisson & Normal Distribution

Uploaded by

Copyright:

Available Formats

Binomial, Poisson & Normal Distribution

(For M.B.A. I Semester)

by Prof. Nitin Karoulia

Theorem 12.6. The number of different combinations of n

determined. For determining the probability in ascending order we use

Here P(r) = probability of ‘r’ time success

Poisson Distribution as a Limiting Form of Binomial Distribution Under the

Distance from Mean (in Ratio of Area from Mean ordinate

You might also like

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.