Scribd
Upload
280 views28 pages

Bernoulli Distribution

The document discusses Bernoulli and binomial distributions. A Bernoulli distribution describes a random variable that can take two values with probabilities p and q=1-p. Bernoulli trials refer to independent experiments with two possible outcomes. The binomial distribution describes the number of successes in n independent Bernoulli trials, where the probability of success is p for each trial. Key properties of the binomial discussed include the mean (np), variance (npq), and moment generating function. Examples are provided of calculating probabilities for binomial random variables.

Uploaded by

Deepak
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
280 views28 pages

Bernoulli Distribution

The document discusses Bernoulli and binomial distributions. A Bernoulli distribution describes a random variable that can take two values with probabilities p and q=1-p. Bernoulli trials refer to independent experiments with two possible outcomes. The binomial distribution describes the number of successes in n independent Bernoulli trials, where the probability of success is p for each trial. Key properties of the binomial discussed include the mean (np), variance (npq), and moment generating function. Examples are provided of calculating probabilities for binomial random variables.

Uploaded by

Deepak
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 28

 A discrete random variable X which takes two

values 0 and 1 with probabilities q and p


respectively. Ie. P(X=1)=p and P(X=0)=q, q=1-p
is called a Bernoulli variate and is said to have a
Bernoulli distribution

 A random experiment which has two outcomes


say happening and non happening(success and
failure)can be repeated in many trials. These
trials are known as Bernoulli trials.
 . Mean (μ)= E[x]
 =∑ x f(x)
 = 1*p+ 0*q
 =p
 Variance (σ2) = E[X2]- E[X]2
 = (12*p+02*q)-p2
 =p- p2
 =p(1-p)
 =pq


 Moment generating function
 MX(t) = E [etx]
 = ∑ etx f(x)
 = et *p+ e0 *q
 = q+p et
 Binomial distribution was discovered by
James Bernoulli in 1700
 Let a random experiment be performed
repeatedly and let the occurrence of an event
in a trial be called success and its non
occurrence a failure. Consider a set of
independent Bernoulli trials in which x
success occur and n-x failure occur. The
probability of getting x success is termed as
binomial distribution
 A random experiment results in 2 mutually
exclusive outcomes termed as success and
failure
 Random experiment repeated in n trials.
Where n is finite. These trials are called
Bernoulli trials
 The trials are independent of each other
 The probability of success p is constant for
each trial
 The probability of failure q= 1-p
 A discrete random variable X is said to follow
binomial distribution if its probability mass
function is given by
f(x) = nCx px q(n-x) , x=0,1....n
=0, else where.
 Mean (μ)= E[x] =∑ x f(x)

 = nCx px q(n-x)

 = px q(n-x)

 =np p(x-1) q(n-x)

 =np (n-1)C(x-1) p(x-1) q(n-x)



 =np (p+q)(n-1) =np* 1 =np


 Variance (σ2) = E[X2]- E[X]2

 = - (np)2

 Put X2 = [x(x-1)+x]

 E[X2] =


 = +
 E[X2] = + np

 = + np

 = n(n-1)p2 + np

 = n(n-1)p2 + np

 = n(n-1)p2 (p+q)(n-2) + np

 E[X2] = n(n-1)p2 (p+q)(n-2) + np
 = n(n-1)p2 *1 + np

 Variance (σ2) =[ n(n-1)p2 *1 + np] -(np)2



 =[ n2p2 - np2 + np] -n2p2

 = -np2 + np

 = np(1-p) =npq
 standard deviation √ npq


 Moment generating function
 MX(t) = E [etx]
 = ∑ etx f(x)

 =

 =

 =
 Mean of Binomial distribution is 3 and
variance is 4 Comment on the distribution
 Ans: Mean =np =3 ------(1)
 Variance = npq =4 ------(2)
 substitute (1) in (2)
 3q = 4
 q =4/3 = 1.3 >0 which
is not possible because q is the probability of
failure. Probability could not exceed 1
 This Binomial distribution does not exist.
 Mean and variance of a distribution are 4 and
4/3. Find P[x≥1]
 Ans: Mean =np =4 ------(1)
 Variance = npq =4/3 ------(2)
 substitute (1) in (2)
 4q = 4/3
 q =1/3, p =1-q =2/3
Apply in (1)
2n/3 =4, n =6

 f(x) = 6Cx (2/3)x (1/3)(6-x) x= 0,1,2,3.....6
 =0, elsewhere
 P[x≥1] = P[ x=1,2,3,4,5,6]
 = 1-P[x=0]
 = 1- [6C0 (2/3)0 (1/3)(6-0)]
 = 1- (1/3)6
 = 1- 1/729
 = 0.9986
 10 coins are thrown simultaneously. Find the
probability of getting at least 7 heads
 Ans: n=10, p=1/2, q=1/2
 f(x) = 10Cx (1/2)x (1/2)(10-x) x=0,1,...10
 =0, elsewhere
 P[at least 7 heads]=P[ x= 7,8,9,10]
 = 10C7 (1/2)7 (1/2)3 + 10C8 (1/2)8 (1/2)2 +
 10C9 (1/2)9 (1/2)1 + 10C10 (1/2)10 (1/2)0
 = 10C (1/2)7 (1/2)3 + 10C (1/2)8 (1/2)2 +
3 2
 10C1 (1/2)9 (1/2)1 + 10C0 (1/2)10 (1/2)0
 P[at least 7 heads]
 = (10*9*8)/3! (1/2)10 + (10*9)/2! (1/2)10
 (10)/1! (1/2)10 + (1/2)10
 =120* (1/2)10 +45* (1/2)10 +10* (1/2)10
 +(1/2)10
 =(1/2)10 [120+45+10+1]
 = (1/2)10 [176]
 =176/1024
 = 0.1718
 The average % of failures in a certain
examination is 40. What is the probability
that out of a group of 6 candidates 1) at least
4 passes in the examination 2) the number
of passes is between 2 and 4
 Ans: n=6,
 probability of a pass =60/100=0.6
probability of a failure =40/100=0.4
 f(x) = 6Cx (0.6)x (0.4)(6-x) x=0,1,...6
 =0, elsewhere
 1. P[at least 4 passes in the examination]
 = P[x= 4,5,6]
 = 6C4 (0.6)4 (0.4)2+6C5 (0.6)5 (0.4)1 +6C6
 (0.6)6 (0.4)0
 =6C2 (0.6)4 (0.4)2+6C1 (0.6)5 (0.4)1 +6C0 (0.6)6
 (0.4)0
 = 15*(0.6)4 (0.4)2+6* (0.6)5 (0.4)1 +(0.6)6 (0.4)0

 = 0.544

 2.
 P[the number of passes is between 2 and 4]
 = P[x= 2,3,4]
 = 6C2 (0.6)2 (0.4)4+6C3 (0.6)3 (0.4)3 +6C4
 (0.6)4 (0.4)2
 = 15(0.6)2 (0.4)4+20(0.6)3 (0.4)3 +15 (0.6)4
 (0.4)2
 = 0.726
 X follows Binomial distribution with
parameters n=6 and p. If 9 P[x=4]=p[x=2]
find p
 Ans: Given 9 P[x=4]=p[x=2]
 9* 6C4 (p)4 (q)2 =6C2 (p)2 (q)4
 9* 6C2 (p)4 (q)2 =6C2 (p)2 (q)4
 9* (p)2 =(q)2
 q =3p
 1-p=3p
 1=4p; p=1/4=0.25
 Let a frequency table be given for which X
takes the values 0,1,2...n and the
corresponding frequencies are f0, f1,....fn
 =N. If it is assumed that X follows the
binomial distribution with parameters n and
p. P[X=x] = f(x) = nCx px q(n-x) and if the
total number of obsrvations is N, we expect
that x no of cases in N* nCx px q(n-x) cases.
So this is called the expected or theortical
frequencies. The process of determining the
most appropriate values of the parameters
 From the given observations and writing
down the pdf is known as fitting of
distribution
 Next we have to determine n and p from the
number of values of X in the data. Let ͞X be
the mean of the data. Mean of the binomial
distribution is np. The estimation of p is usually
done by equating ͞X and np
 np= ͞X p= ͞X/n
Then fit the pdf as
f(x) = nCx px q(n-x) , x=0,1....n
=0, else where
And the theoretical frequencies are N*f(0),
N*f(1).......N*f(n) respectively.
 The following table gives the number of
heads in 103 repetitions of 4bthrows of a
biased coin. Fit an appropriate binomial
distribution an calculate the theoretical
frequencies
 No. of heads: 0 1 2 3 4 Total
 Frequency : 8 32 34 24 5 103
 Ans: np= ͞X
 where ͞X =
X f Xf
0 8 0
1 32 32

2 34 68
3 24 72
4 5 20
Total 103 192
 ͞X = 192/103= 1.86
 p= 1.86/4 =0.46
 q= 0.54

f(x) = 4Cx 0.46x 0.54(4-x) , x=0,1....4


=0, else where
Theoretical frequencies are N*f(0), N*f(1).......N*f(n)
 where N = 103
x f(x) Theortical
frquency N*f(x)
0 = 4C0 0.460 0.544 N*f(0)= 8.8 ~9
1 = 4C1 0.461 0.543 N*f(1)= 29.8 ~30

2 = 4C2 0.462 0.542 N*f(2)= 38.1~38

3 = 4C3 0.463 0.541 N*f(3)= 21.7 ~22

4 = 4C4 0.464 0.540 N*f(4)= 4.5 ~4

Total 103

You might also like