Module – 4 Probability distributions

School of Advanced Sciences

Department of Mathematics 1
2
3
Binomial Distribution

Constants of Binomial Distribution

4
5
6
7
8
9
10
11
Note:
Poisson distribution is a limiting case of
binomial distribution under the following
assumptions:
The number of trials 𝑛is infinitely large,
i.e., 𝑛 → ∞.
The probability of success for each trial is
infinitely small, i.e., 𝑝 → 0.
𝑛𝑝 = 𝜆is finite.

12
13
Solution
Given,

Since 𝑛 is large and 𝑝 is small, Poisson distribution is

used.

The Poisson probability distribution is

(i) Probability that there are 2 defectives

14
A firm owns two cars and hires one every day. The number of demands for a car on each day is distributed as a
Poisson distribution with mean 1.5. Calculate the proportion of days
(i) on which there is no demand
(ii) on which demand is refused.
Solution P(X>2)=1−P(X≤2)
For a Poisson distribution, =1− 𝑃 𝑋 =0 +𝑃 𝑋 =1 +𝑃 𝑋 =2
𝑒 −𝜆 𝜆𝑥 = 0.192
𝑃 𝑋=𝑥 =
𝑥! Therefore, the number of days in a year on which demand is
Given, refused is
𝜆 = 1.5 365 × 0.192 ≈ 70 days
(i) Proportion of days on which there is no
demand
𝑒 −1.5 ቀ1.5)0
𝑃 𝑋=0 =
0!
= 𝑒 −1.5= 0.223
Hence, the proportion of days in a year on
which there is no demand for a car is
365 × 0.223 ≈ 81 days
(ii) Proportion of days on which demand is
refused
Demand is refused only when the demand
exceeds the available cars, i.e., when 15
𝑋>2
16
17
18
19
20
Properties of Normal Distribution
• Mean = 𝜇= Median = Mode
• Variance = 𝜎 2

21
22
Standard Normal Distribution
The normal distribution 𝑁 0 1 is called the standard normal distribution (SND).
If a random variable 𝑋follows a normal distribution with mean 𝜇and variance 𝜎 2 ,its transformation
𝑋−𝜇
𝑍=
𝜎
follows the standard normal distribution
mean = 0 and 𝜎 2 = 1

The density function

The distribution function of the SND

23
24
25
Problem
The mean and standard deviation of a normal variate are 8 and 4 respectively. Find
(i) 𝑷 𝟓 ≤ 𝑿 ≤ 𝟏𝟎
(ii) 𝑷 𝑿 ≥ 𝟓 (ii) 𝑃 𝑋 ≥ 5
𝑃 𝑋 ≥ 5 = 𝑃 𝑍 ≥ −0.75
Solution = 1 − 𝑃 𝑍 ≤ −0.75
Given, = 1 − 𝐹 −0.75
𝜇 = 8, 𝜎 = 4 = 1 − 0.2266 = 0.7734
Standard normal transformation:
𝑋−𝜇 𝑋−8
𝑍= =
𝜎 4
(i) 𝑃 5 ≤ 𝑋 ≤ 10
When 𝑋 = 5,
5−8
𝑍= = −0.75
4
When 𝑋 = 10,
10 − 8
𝑍= = 0.5
4
Hence,
𝑃 5 ≤ 𝑋 ≤ 10 = 𝑃 −0.75 ≤ 𝑍 ≤ 0.5
= 𝐹 0.5 − 𝐹 −0.75
From standard normal table,
𝐹 0.5 = 0.6915, 𝐹 −0.75 = 0.2266
𝑃 5 ≤ 𝑋 ≤ 10 = 0.6915 − 0.2266 = 0.4649 26
In a sample of 1000 cases, the mean of a certain test is 14 and the standard deviation is 2.5. Assuming the distribution to be
normal, find
(i) how many students score between 12 and 15
(ii) how many score above 18
(iii) how many score below 18
Solution
Given,
𝑁 = 1000, 𝜇 = 14, 𝜎 = 2.5
Standard normal transformation:
𝑋 − 𝜇 𝑋 − 14
𝑍= =
𝜎 2.5
(i) Number of students scoring between 12 and 15
𝑃 12 ≤ 𝑋 ≤ 15
When 𝑋 = 12,
12 − 14
𝑍= = −0.8
2.5
When 𝑋 = 15,
15 − 14
𝑍= = 0.4
2.5
𝑃 12 ≤ 𝑋 ≤ 15 = 𝑃 −0.8 < 𝑍 < 0.4
= 𝐹 0.4 − 𝐹 −0.8
From standard normal table,
𝐹 0.4 = 0.6554, 𝐹 −0.8 = 0.2119
𝑃 12 ≤ 𝑋 ≤ 15 = 0.6554 − 0.2119 = 0.4435
27
Number of students = 1000 × 0.4435 ≈ 444
(ii) Number of students scoring above 18
𝑃 𝑋 > 18
18 − 14
𝑍= = 1.6
2.5
𝑃 𝑋 > 18 = 𝑃 𝑍 > 1.6
= 1 − 𝑃 𝑍 ≤ 1.6 = 1 − 𝐹 1.6
From table,
𝐹 1.6 = 0.9452
𝑃 𝑋 > 18 = 1 − 0.9452 = 0.0548
Number of students = 1000 × 0.0548 ≈ 55

(iii) Number of students scoring below 18

𝑃 𝑋 < 18
𝑃 𝑋 < 18 = 𝑃 𝑍 < 1.6 = 𝐹 1.6
= 0.9452
Number of students = 1000 × 0.9452 ≈ 945

28
29
30
31
Question
The local corporation authorities in a certain city install 10,000 electric lamps in the streets of the city with the
assumption that the life of the lamps is normally distributed. If these lamps have an average life of 1,000 burning
hours with a standard deviation of 200 hours, calculate
(i) the number of lamps expected to fail in the first 800 hours, and
(ii) the number of lamps expected to fail between 800 and 1,200 hours.
Ans: (i) 1587 lamps
(ii) 6826 lamps

32
Gamma Distributions (Continuous)
The Gamma integral is defined as
∞
Γ 𝑛 = න 𝑥 𝑛−1 𝑒 −𝑥 𝑑𝑥.
0

33
34
35
1) In a certain city, the daily consumption of electric power (in millions of kilowatt-hours) can be treated as a
random variable having an Erlang distribution with parameters
𝟏
𝝀 = , 𝒌 = 𝟑.
𝟐
If the power plant has a daily capacity of 12 million kWh, find the probability that the power supply will be
inadequate on any day.
Solution
Let 𝑋 be the daily consumption (in million kWh).
1
Since 𝑋 ∼ Erlang 𝑘 = 3𝜆 = 2 ,its pdf is
𝜆𝑘
𝑓 𝑥 = 𝑥 𝑘−1 𝑒 −𝜆𝑥 , 𝑥 > 0.
𝑘−1 !
1
So with 𝑘 = 3, 𝜆 = 2:
ቀ1/2)3 1 2 −𝑥/2
𝑓 𝑥 = 𝑥 2 𝑒 −𝑥/2
= 𝑥 𝑒 , 𝑥 > 0.
2! 16
“Inadequate supply” means consumption exceeds capacity:
𝑃 inadequate = 𝑃 𝑋 > 12 .
For an Erlang with integer 𝑘, the survival function is
ቀ𝜆𝑥)2 ቀ𝜆𝑥)𝑘−1
𝑃 𝑋 > 𝑥 = 𝑒 −𝜆𝑥 1 + 𝜆𝑥 + + ⋯+ .
2! 𝑘−1 !

36
Here 𝑘 = 3, so
ቀ𝜆𝑥)2
𝑃 𝑋 > 𝑥 = 𝑒 −𝜆𝑥 1 + 𝜆𝑥 + .
2
1
Now substitute 𝜆 = 2and 𝑥 = 12:
1
𝜆𝑥 = ⋅ 12 = 6
2
62
𝑃 𝑋 > 12 = 𝑒 −6 1+6+ = 𝑒 −6 1 + 6 + 18 = 25𝑒 −6 .
2
Numerically,
25𝑒 −6 ≈ 0.06197 ≈ 0.0620.
𝑃 power supply inadequate = 𝑃 𝑋 > 12 = 25𝑒 −6 ≈ 0.062

37
 2) If a company employs 𝒏salespersons, its gross sales (in thousands of rupees) may be regarded as a random
𝟏
variable having an Erlang distribution with parameters 𝝀 = 𝟐and 𝒌 = 𝟖𝟎 𝒏 .If the sales cost is Rs. 8000 per
salesperson, how many salespersons should the company employ in order to maximise the expected profit?
Solution:
Let 𝑋 represent the gross sales (in rupees) by 𝑛sales persons.
𝑋 follows the Erlang distribution with parameters
1
𝜆 = 2and 𝑘 = 80,000 𝑛.
𝑘
∴ 𝐸 𝑋 = = 1,60,000 𝑛
𝜆
If 𝑦 denotes the total expected profit of the company, then
𝑦 = total expected sales − total sales cost
𝑦 = 1,60,000 𝑛 − 8000𝑛

= 0, when 𝑛 = 10 or 𝑛 = 100
𝑑2𝑦 40,000
= − 3 < 0, when 𝑛 = 100.
𝑑𝑛2
𝑛2
Therefore, 𝑦 is maximum, when 𝑛 = 100.
That is the company should employ 100 salespersons in order to maximise the total expected profit.

38
3) Consumer demand for milk in a certain locality, per month, is known to be a general Gamma (Erlang) RV. If the
average demand is 𝒂 liters and the most likely demand is 𝒃 liters 𝒃 < 𝒂 ,what is the variance of the demand?
Solution: Let 𝑋 represent the monthly consumer demand of milk.

Average demand is the value of 𝐸 𝑋 .

Most likely demand is the value of the mode of 𝑋 or the value of 𝑋 for which its density function is maximum.
If 𝑓 𝑥 is the density function of 𝑋, then
𝜆𝑘 𝑘−1 −𝜆𝑥
𝑓 𝑥 = 𝑥 𝑒 ,𝑥 > 0
Γ 𝑘

𝑘−1 𝑘
So, mode = = 𝑏 and mean = 𝜆 = 𝑎.
𝜆

39
40
Exponential Distributions
A random variable 𝑋is said to possess an exponential distribution if its density can be expressed in the form
−𝜆𝑥
𝑓 𝑥 = ቊ𝜆𝑒 , 𝑥 ≥ 0,
0, otherwise.
Properties
1
•Mean = 𝜆
1
•Variance = 𝜆2
𝑡 −1
•MGF: 𝑀𝑋 𝑡 = 1 − 𝜆
Distribution of the Exponential distribution

Memoryless property of the Exponential distribution

➢ The exponential probability distribution is useful in describing the time it takes to complete a task.
➢ The exponential random variables can be used to describe:
•Time between vehicle arrivals at a toll booth
•Time required to complete a questionnaire
•Distance between major defects in a highway
➢ In waiting line applications, the exponential distribution is often used for service times.

41
42
43
44
45
46
47
 (b) 𝑃 1.5 ≤ 𝑋 ≤ 3
𝑃 1.5 ≤ 𝑋 ≤ 3 = 𝑃 𝑋 ≤ 3 − 𝑃 𝑋 < 1.5
= 1 − 𝑒 −3/3 − 1 − 𝑒 −1.5/3
= 𝑒 −0.5 − 𝑒 −1
= 0.6065 − 0.3679 = 0.2386
(c) 𝑃 𝑋 > 3.5 ∣ 𝑋 > 2.5
Using memoryless property:
𝑃 𝑋 > 3.5 ∣ 𝑋 > 2.5 = 𝑃 𝑋 > 3.5 − 2.5 = 𝑃 𝑋 > 1
Given = 𝑒 −1/3 = 0.7165
𝑋= life time of an electric bulb (in thousands of hours).
1
𝑋is exponential with rate 𝜆 = .
3
So the pdf is
1
𝑓 𝑥 = 𝜆𝑒 −𝜆𝑥 = 𝑒 −𝑥/3 , 𝑥 > 0
3
And the CDF form we use:
𝑃 𝑋 > 𝑡 = 𝑒 −𝜆𝑡 = 𝑒 −𝑡/3
(a) 𝑃 𝑋 > 3
𝑃 𝑋 > 3 = 𝑒 −3/3 = 𝑒 −1 = 0.3679

48
 𝟏
2. The time (in hours) required to repair a machine is exponentially distributed with parameter 𝝀 = 𝟐.
(a) What is the probability that the repair time exceeds 2 hours?
(b) What is the conditional probability that a repair takes at least 10 hours, given that its duration exceeds 9 hours?
Given
Let 𝑋be the repair time (hours).
1
𝑋 ∼ Exponential 𝜆 =
2
1
𝑓 𝑥 = 𝜆𝑒 −𝜆𝑥 = 𝑒 −𝑥/2 , 𝑥 > 0
2
−𝜆𝑡
𝑃 𝑋>𝑡 =𝑒 = 𝑒 −𝑡/2 .
(a) Probability that repair time exceeds 2 hours: 𝑃 𝑋 > 2

𝑃 𝑋 > 2 = 𝑒 −2/2 = 𝑒 −1 = 0.3679.

(b) Conditional probability 𝑃 𝑋 ≥ 10 ∣ 𝑋 > 9

Using the memoryless property:

𝑃 𝑋 ≥ 10 ∣ 𝑋 > 9 = 𝑃 𝑋 > 10 ∣ 𝑋 > 9 = 𝑃 𝑋 > 10 − 9 = 𝑃 𝑋 > 1 .

𝑃 𝑋 > 1 = 𝑒 −1/2 = 0.6065.
49
Exercise
1. The mileage which car owners get with certain kind of radial tyre is a random variable having an exponential distribution with
mean 4,000 km. Find the probabilities that one of these tires will last (i) at least 2,000 km (ii) at most 3,000 km.
Solution: (i) 0.6065 (ii) 0.5270

2. The amount of time that a camera will run without having to be reset is a random variable having exponential distribution with
mean 50 days. Find the probability that such a camera will (i) have to reset in less than 20 days (ii) not have to be reset in atleast
60 days.
Solution: (i) 0.3297 (ii) 0.3012

3. Assume that the length of phone calls made at a particular telephone booth is exponentially distributed with mean of 3 minutes.
If you arrive at the telephone booth just Chris about to make a call, find the following (i) the probability that you will wait
more than 5 minutes before Chris is done with the call (ii) the probability that chris call will last between 2 minutes and 6
minutes.
Solution: (i) 0.1889 (ii) 0.3781

4. Suppose the life of automobile batteries is exponentially distributed with parameter 𝜆 = 0.0001 days. (a) What is the
probability that a battery will last more than 1200 days? (b) What is the probability that a battery will last more than 1200 days
given that it has already served 1000 days?
Solution: (i) 0.3012 (ii) 0.8187

50
Weibull distribution

51
52
53
54
 Exercise

1. The life time of a component measured in hours follows Weibull distribution with parameter 𝛼 = 0.2, 𝛽 = 0.5. Find the
mean life time of the component.
Solution: 50

2. Suppose that the service life (in hours) of a semi-conductor is a random variable having the weibull distribution with
parameter 𝛼 = 0.025and 𝛽 = 0.5. What is the probability that such a semi-conductor will be in working condition after
4000 hours?
Solution: 0.2057
3. Suppose there are twelve multiple choice questions in an English class quiz. Each question has five possible answers and only
one is correct. Find the probability of having four or fewer correct answers if a student attempts to answer every question at
random.
Solution: 0.9274
4. If 10% of the screws produced by an automatic machine are defective, find the probability that out of 20 screws selected at
random, there are
(i) exactly 2 defective
(ii) at least 2 defective
(iii) between 1 and 3 defective (inclusive).
Solution: (i) 𝑃 𝑋 = 2 = 0.2852
(ii) 𝑃 𝑋 ≥ 2 = 0.6083
(iii) 𝑃 1 ≤ 𝑋 ≤ 3 = 0.7344
55
5. For a Poisson distribution with 𝜆 = 12.33, find
(i) 𝑃 𝑋 ≤ 15
(ii) 𝑃 𝑋 = 6
(iii) 𝑃 𝑋 > 15
(iv) 𝑃 𝑋 ≥ 15
(v) 𝑃 8 ≤ 𝑋 ≤ 10 .
Solution: (i) 𝑃 𝑋 ≤ 15 = 0.8196
(ii) 𝑃 𝑋 = 6 = 0.0216
(iii) 𝑃 𝑋 > 15 = 0.1804
(iv) 𝑃 𝑋 ≥ 15 = 0.2586
(v) 𝑃 8 ≤ 𝑋 ≤ 10 = 0.2376
6. In a photographic process the developing times of prints may be treated as a normal random variable with 𝜇 = 16.28 and 𝜎
= 0.12sec. Find the probability that it will take
(i) at least 16.20 seconds
(ii) at most 16.35 seconds to develop one print.
Solution: (i) 𝑃 𝑋 ≥ 16.20 = 0.7475
(ii) 𝑃 𝑋 ≤ 16.35 = 0.7202
7. The weekly wages of 1000 workers are normally distributed around a mean of Rs 70 with S.D. Rs 5. Estimate the number of workers
whose weekly wages will be
(i) between Rs 69 and Rs 72
(ii) less than Rs 69
(iii) more than Rs 72.
Solution: (i) 1000 × 𝑃 69 ≤ 𝑋 ≤ 72 ≈ 235 workers
(ii) 1000 × 𝑃 𝑋 < 69 ≈ 421 workers
(iii) 1000 × 𝑃 𝑋 > 72 ≈ 345 workers 56