Scribd
Upload
20 views10 pages

Understanding Geometric Distribution Concepts

The document explains the geometric distribution, which models the number of trials needed to achieve the first success in a series of independent Bernoulli trials. It outlines the properties, probability function, mean, variance, moment generating function, and the memoryless property of the distribution. Additionally, it provides several examples demonstrating its application in real-world scenarios.

Uploaded by

Muhammad Saad
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
20 views10 pages

Understanding Geometric Distribution Concepts

The document explains the geometric distribution, which models the number of trials needed to achieve the first success in a series of independent Bernoulli trials. It outlines the properties, probability function, mean, variance, moment generating function, and the memoryless property of the distribution. Additionally, it provides several examples demonstrating its application in real-world scenarios.

Uploaded by

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

Geometric Distribution

Geometric Experiment

Geometric experiment possesses the following properties.

i. The outcomes of each trail may be classified into one of two categories, i.e. success &
failure.
ii. The probability of success is denoted by p, remains constant.
iii. The successive trails are independent.
iv. The experiment is repeated a variable number of times until first success appear.

Explanation: Let X trails are performed to obtain first success.

Geometric distribution

Let X is number of trails needed for first success is to appear, then X is a geometric random
variable and its probability function is given by;
X −1
f ( x )=q p X=1 , 2 ,3 , … .

Remember that f (x) at x=0 is not possible.

Properties:

i. f ( x ) ≥ 0 for all X .
ii. ∑ f ( x )=1
Proof:
∞

∑ f ( x )= ∑ q X −1 p
X=1
∞

∑ f ( x )= p ∑ q X −1
X =1

∑ f ( x )= p [ q0 + q1 +q 2+⋯ ]
∑ f ( x )= p [ 1+ q+q 2 +⋯ ]
∑ f ( x )= p ( 1−q )−1
∑ f ( x )= p p−1
∑ f ( x )=1
Mean and Variance
∞
E ( X )=∑ Xf (x ) E ( X )= ∑ X q X −1 p
X=1

∞
E ( X )=p ∑ X q X −1
X =1

E ( X )=p [1. q 1−1+2. q 2−1+3. q 3−1+ 4. q4 −1+⋯ ]

E ( X )=p [1+2. q 1+ 3.q 2 +4. q 3+⋯ ]


−2
E ( X )=p (1−q )

E ( X )=p p−2

1
E ( X )=
p

Note: E( X )≥ 1

E ( X 2) =E [ X ( X −1 )+ X ]

E ( X 2) =E ( X ( X −1 ) ) + E ( X )

1
E ( X ) =E ( X ( X −1 ) ) + →(1)
2
p

E ( X ( X −1 ) )=∑ ( X ( X−1 ) ) f (x )
∞
E ( X ( X −1 ) )= ∑ ( X ( X −1 ) ) q X −1 p
X =1
∞
E ( X ( X −1 ) )= p ∑ ( X ( X−1 ) ) q X −1
X =1

E ( X ( X −1 ) )= p [ 2.1 q 2−1 +3.2 q3 −1 + 4.3 q4 −1+5.4 q5−1 +⋯ ]

E ( X ( X −1 ) )= p [ 2.1 q+3.2 q 2+ 4.3 q 3+5.4 q 4 +⋯ ]

E ( X ( X −1 ) )= p 2 q [ 1+3. q+ 6 q2 +10. q3 +⋯ ]

E ( X ( X −1 ) )=2 qp ( 1−q )
−3

E ( X ( X −1 ) )=2 qp p
−3

2q
E ( X ( X −1 ) )= 2
p

Substitute in equation 1.

1
E ( X ) =E ( X ( X −1 ) ) + →(1)
2
p

2q 1
E ( X )=
2
+
p p
2

Var ( X )=E ( X 2 )−( E (X ) )


2

()
2
2q 1 1
Var ( X )= 2
+ −
p p p

2q 1 1
Var ( X )= + − 2
p p p
2

2 q+ p−1
Var ( X )= 2
p

2 q+ p−( q+ p)
Var ( X )=
p2

2 q+ p−q− p
Var ( X )= 2
p

q
Var ( X )= 2
p

Moment Generating Function of Geometric Distribution


M 0 ( t )=E ( etx )

M 0 ( t )=∑ e q p
tx x

M 0 ( t )= p ∑ ( q e )
t x

M 0 ( t )= p [ ( q e t ) + ( q e t ) + ( q et ) +⋯ ]
0 1 2

M 0 ( t )= p [ 1+ ( q e t ) + ( q et ) +⋯ ]
1 2

t −1
M 0 ( t )= p ( 1−q e )

p
M 0 ( t )=
( 1−q e t )
Moments from m.g.f

¿ d p
μ1 =
dt ( 1−q et )

( 1−q e t ) d p− p d ( 1−q e t )
dt dt
μ¿1= t 2
( 1−q e )

¿ pq ( e t )
μ1 = 2
→1
( 1−q e t )
Put t=0

¿ pq q
μ1 = =
( 1−q ) p
2

¿ d ¿
μ2 = M (t)
dt 0

d pq ( e )
t
μ¿2=
dt ( 1−q et )2

( 1−q e t ) d ( pq et )−( pq e t ) d
2

¿
dt dt ( 1−q e )
t 2
μ=
2 4
( 1−q e t )
2
¿ ( 1−q e t ) ( pq et )−( pq et ) 2 ( 1−q e t ) (−q et )
μ=
2 4
( 1−q et )
Put t = 0

( 1−q )2 ( pq )+ ( pq ) 2 ( 1−q ) ( q )
μ¿2= 4
( 1−q )
3 2 2
¿ p q+2 p q
μ2 = 4
p
2
¿q 2q
μ= + 2
2
p p

Memoryless property

The probability function of geometric distribution


X −1
f ( x )=q p X=1 , 2 ,3 , … .

P ( X >k )=P ( X=k +1 ) + P(X =k +2)+ P(X =k +3)+⋯


k k +1 k+ 2
P ( X >k )=q p+ q p+q p+⋯

P ( X >k )=q k p ( 1+q 1 +q2 +⋯ )


k −1
P ( X >k )=q p ( 1−q )
k
P ( X >k )=q

P ( X >s +t ∩ X >s )
P ( X > s+t ⋰ X > s )=
P(X > t)

P ( X >s +t )
P ( X > s+t ⋰ X > s )=
P( X >t)
s+t
q
P ( X > s+t ⋰ X > s )= t
q
s
P ( X > s+t ⋰ X > s )=q

Example:
A safety engineer feels that 35% of all industrial accidents in her plant are caused by failure of
employees to follow instructions. She decides to look at the accident reports (selected randomly
and replaced in the file after reading) until she finds one that shows an accident caused by failure
of employees to follow instructions. On average, how many reports would the safety
engineer expect to look at until she finds a report showing an accident caused by employee
failure to follow instructions? What is the probability that the safety engineer will have to
examine at least three reports until she finds a report showing an accident caused by employee
failure to follow instructions?

Solution:

Let X= the number caused by employee failure to follow instructions. X takes on the values 1, 2,
3, …. The first question asks you to find the expected value or the mean. The second question
asks you to find P(x≥3).

p=0.35

q=1− p=0.65

1 1
E ( X )= = =2.857
p 0.35

P ( X ≥3 )=1−P( X <3)
x−1
f ( x )=q p
2
P ( X <3 )= ∑ ( 0.65 )
X −1
( 0.35 )
X=1

2
P ( X <3 )=( 0.35 ) ∑ ( 0.65 )
X −1

X=1

P ( X <3 )=( 0.35 ) [ ( 0.65 )1−1 + ( 0.65 )2−1 ]

P ( X <3 )=( 0.35 ) [ ( 0.65 )0 + ( 0.65 )1=0.5775 ]

P ( X ≥3 )=1−P( X <3)

P ( X ≥3 )=1−0.5775=0.4225

Example:

An instructor feels that 15% of students get below a C on their final exam. She decides to look
at final exams (selected randomly and replaced in the file after reading) until she finds one that
shows a grade below a C. We want to know the probability that the instructor will have to
examine at least ten exams until she finds one with a grade below a C. What is the probability
question stated mathematically?

Solution:

p=0.15

q=1− p=0.85

P ( X ≥10 )=?

P ( X ≥10 )=1−P( X <10)


x−1
f ( x )=q p
9
P ( X <10 ) = ∑ ( 0.85 )
X −1
( 0.15 )
X =1

9
P ( X <10 ) =( 0.15 ) ∑ ( 0.85 )
X −1

X =1

P ( X <10 ) =( 0.15 ) [ ( 0.85 )1−1 + ( 0.85 )2−1 + ( 0.85 )3−1 + ( 0.85 )4 −1+ ( 0.85 )5−1+ ( 0.85 )6−1 + ( 0.85 )7−1 + ( 0.85 )8−1 + ( 0.85 )9−1 ]

P ( X <10 ) =( 0.15 ) [ ( 0.85 )0 + ( 0.85 )1+ ( 0.85 )2+ ( 0.85 )3 + ( 0.85 )4 + ( 0.85 )5 + ( 0.85 )6 + ( 0.85 )7+ ( 0.85 )8 ]

P ( X <10 ) =( 0.15 ) [ 4.307 ]=0.646

P ( X ≥10 )=1−0.646=0.354

Example:
Suppose that you are looking for a student at your college who lives within five miles of you.
You know that 55% of the 25,000 students do live within five miles of you. You randomly
contact students from the college until one says he or she lives within five miles of you. What is
the probability that you need to contact four people?
a. Let X= the number of ____________ you must ask ____________ one says yes.
b. What values does X take on?
c. What are p and q?
d. The probability question is P(_______).

Solution: This is a geometric problem because you may have a number of failures before you
have the one success you desire. Also, the probability of a success stays the same each time you
ask a student if he or she lives within five miles of you. There is no definite number of trials
(number of times you ask a student).
a. Let X= the number of students you must ask until one says yes.
b. X =1, 2, 3, ……, 2500
c. P = 0.55 and q = 0.45
d. The probability question is P( X = 4)
x−1
f ( x )=q p

3
P ( X=4 )=( 0.45 ) ( 0.55 )=0.075

Example:
Assume that the probability of a defective computer component is 0.02. Components are
randomly selected. Find the probability that the first defect is caused by the seventh component
tested. How many components do you expect to test until one is found to be defective?
Solution:
Let X= the number of computer components tested until the first defect is found.
X takes on the values 1, 2, 3, … where p=0.02.
q=1− p=0.98
6
P ( X=7 )=( 0.98 ) ( 0.02 )=0.0177

The expected number of defective components:

1 1
E ( X )= = =50
p 0.02

The variance of defective components:

0.98
Var ( X )= 2
=2450
(0.02)

Example:

The lifetime risk of developing pancreatic cancer is about one in 78 (1.28%). Let X= the number
of people you ask until one says he or she has pancreatic cancer. Then X is a discrete random
variable with a geometric distribution:

a. What is the probability of that you ask ten people before one says he or she has pancreatic
cancer?
b. What is the probability that you must ask 20 people?
c. Find the (i) mean and (ii) standard deviation of X.
Solution:

1
p= =0.0128
78

q=1− p=0.9872
a. P ( X=10 )=?
9
P ( X=10 )=( 0.9872 ) ( 0.0128 ) =0.0114
b. P ( X=20 )=?
19
P ( X=20 )=( 0.9872 ) ( 0.0128 )=0.01002
c. Mean
1 1
E ( X )= = =78.125
p 0.0128
Standard deviation

q
Var ( X )= 2
p

S . D ( X )=
√ q
p2

S . D ( X )=
√ 0.9872
( 0.0128 )
2
=77.62

MCQ’s on Geometric Distribution

1. The mean and Variance of geometric distribution are


i. p/q and p/q
ii. q/p and q/p
iii. q/p and q/p2
iv. p/q and p2/q
2. If the probability of success is 0.5, then the mean of geometric distribution is
i. 1
ii. 2
iii. 5
iv. 7
3. If the probability of success is 0.5, then the variance of geometric distribution is
i. 2
ii. 4
iii. 6
iv. 7
4. In geometric distribution, the last trail is
i. Success
ii. Failure
iii. Both i & ii are possible
iv. Anything possible
5. geometric distribution is of continuous type
a. true
b. false

Common questions

Powered by AI

The moment generating function (mgf) of a geometric distribution, M(t) = pe^(t) / (1-qe^(t)), provides insights into its statistical properties by facilitating the computation of moments. The mgf offers a comprehensive analytic tool for deriving the mean, variance, and higher moments. Key computations using the mgf involve differentiating M(t) with respect to t and evaluating these derivatives at t = 0 to obtain moments. For instance, the first derivative gives the mean, while the second derivative gives the second moment, which aids in variance calculation. This process highlights the distribution's skewness and kurtosis, reflecting its likelihood spread and peak characteristics .

The geometric distribution is highly useful in industries like telecommunications for modeling call drops, manufacturing for defect detection, and logistics for estimating customer arrivals. It offers straightforward calculations for expected values and probabilities due to its memoryless property and simplicity in predicting the number of trials for the first success. However, its limitations include the assumption of constant success probability and independent trials, which may not hold in real-world scenarios where external factors influence outcomes. Additionally, it does not account for sequences of successes and fails, focusing instead solely on the first success . This can limit its applicability in complex models requiring detailed event sequences or dependent trials.

To calculate the probability of an event occurring after a specific number of trials in a geometric setting, one must employ the probability function P(X=k) = (1-p)^(k-1) * p. The steps include determining the probability of success in each trial (p), identifying the trial number of interest (k), and applying the probability function. This relates to practical scenarios such as determining the likelihood that a salesperson makes a sale on the nth call or a machinery defect is found after k tests. For instance, if p = 0.15 for a success on each trial, the probability of a success on the 10th trial would be P(X=10) = (0.85)^9 * 0.15 = 0.0232959, showing the probability of requiring precisely nine failures before achieving success .

In a quality control process, a geometric distribution can be applied to model the number of inspections needed before finding a defective product. For instance, suppose the probability of a component being defective is 0.02. The geometric distribution will determine how many components must be tested before a defect is found. Here, the number of trials X follows a geometric distribution with p = 0.02. The expected number of components tested (mean) is E(X) = 1/p = 50, indicating that one can expect to test 50 components before finding a defect. Additionally, the probability that the first defect occurs at the seventh component is P(X=7) = (0.98)^6 * 0.02 = 0.0177, illustrating the application of these calculations in predicting quality control outcomes .

The memoryless property of a geometric distribution implies that the probability of future occurrences is independent of past events. This property stems from the independence of trials and the constant probability of success. It is formally exhibited as P(X > s + t | X > s) = P(X > t), indicating that the probability of an event happening after s trials is the same as it occurring after zero trials, given it hasn't happened by trial s. This is tested using the probability function P(X > k) = q^k, where q = 1-p and k is the number of trials minus 1. This calculation shows the future probability is not affected by previously occurred trials, thus demonstrating the memoryless characteristic .

The example involving a search for a student living within five miles illustrates the practical use of geometric distribution in population sampling. With a success probability of p = 0.55, the geometric distribution calculates the probability that a specific number of students need to be contacted before one confirms residency within five miles. The probability of needing four contacts is given by P(X=4) = (0.45)^3 * 0.55 = 0.075 . This situation showcases how geometric distribution can model scenarios of repeated random sampling from a large population until a particular characteristic is observed. The results reflect the proportion of the population exhibiting the trait and the efficiency of targeted sampling strategies.

An instructor can utilize geometric distribution to understand trends in student performance by modeling the number of final exams reviewed until one is below a C grade threshold. Assuming a probability of 0.15 for any exam to have a grade lower than a C, the geometric distribution estimates expected review numbers and trial-based probabilities. For no grade below C after nine exams, the probability is P(X>=10) = 1 - P(X<10) = 0.354 . This application helps track performance tendencies, adapt teaching strategies, and allocate resources effectively. By modeling the grade distribution using geometric principles, the instructor gains insights into academic support needs and the randomness of poor performance incidence.

Utilizing geometric distribution to model rare events, such as the pancreatic cancer occurrence with p = 0.0128, underscores its efficacy in predicting the frequency and variation in event observations. In this example, the distribution calculates the probability that exactly ten people are asked before encountering a case of pancreatic cancer, P(X=10) = (0.9872)^9 * 0.0128 = 0.0114 . This provides substantial data interpretation guidance by highlighting the infrequency and unpredictability of rare events. Additionally, the distribution calculates the expected mean and standard deviation, indicating expected intervals and variability in events. These insights are crucial for planning and resource allocation in healthcare and epidemiological studies by conveying the sparseness and challenge in tracking rare conditions.

The mean of a geometric distribution reflects the expected number of trials required to achieve the first success. It is derived by evaluating the sum of the probability function for each possible trial outcome. In the geometric distribution, the mean is calculated using the formula E(X) = 1/p, where p represents the probability of success in each trial . This formula stems from the fact that each trial is independent and has the same probability of success, leading to an accumulation of probabilities that factor in the likelihood of success occurring at different intervals. The implications of this calculation are crucial for understanding scenarios like customer arrivals, defect detections, or other processes where a single occurrence is observed over repeated trials.

In a geometric distribution, the mean (E(X) = 1/p) and variance (Var(X) = q/p^2, where q = 1-p) are inversely related to the probability of success. As the probability of success (p) increases, the mean decreases, indicating fewer expected trials are needed for a success, while the variance also decreases, reflecting lesser variability in the number of trials required for success . Conversely, a lower probability of success results in both a higher mean and a higher variance, indicating more trials are expected for success and increased variability in trial outcomes. This relationship reveals how the predictability of an event changes with the probability of success, highlighting the flexibility of the geometric distribution in modeling situations based on their success rates.

You might also like