Module 4 – Probability Distributions | NA

MAT2001 Statistics for Engineers

Syllabus: Binomial and Poisson distributions – Normal distribution – Gamma distribution –

Exponential distribution – Weibull distribution.

Continuous Probability Distributions

4. Gamma distribution

Gamma distribution for one parameter: The continuous random variable X is said to
follow gamma distribution with a parameter 𝜆, (i.e.,) 𝑋~𝐺(𝜆) if its probability density function
𝑒 −𝑥 𝑥 𝜆−1 ∞
is given by 𝑓(𝑥) = , ∀ 𝜆 > 0, 0 < 𝑥 < ∞, where Γ(𝜆) = ∫0 𝑥 𝜆−1 𝑒 −𝑥 𝑑𝑥 = (𝜆 − 1)!,
Γ(𝜆)
𝜆 > 0 is a gamma function.

Note:
1) MGF = (1 − 𝑡)−𝜆
2) Mean = Variance = 𝜆
3) Γ(𝜆) = (𝜆 − 1) Γ(𝜆 − 1)
1
4) Γ (2) = √𝜋
5) Γ(1) = 1
6) Γ(λ + 1) = λ Γ(λ) = 𝜆!
1
7) Γ (− 2) = −2√𝜋

Generalized form of the gamma distribution (or) Erlang distribution (or)

Gamma distribution with two parameters: A continuous random variable X is said
to follow gamma distribution with the parameters ‘𝑎’ and ‘𝜆’, (i.e.,) 𝑋~𝐺(𝑎, 𝜆) if its probability
𝑎𝜆 𝑒 −𝑎𝑥 𝑥 𝜆−1
density function is given by 𝑓(𝑥) = , ∀ 𝜆 > 0, 𝑎 > 0, 0 < 𝑥 < ∞.
Γ(𝜆)

Application of Gamma distribution:

(i) The amount of rainfall accumulated in a reservoir.
(ii) The size of loan defaults or aggregate insurance claim.
(iii) The load on web servers.
(iv) The many and varied forms of telecom exchange.
(v) The following real-life applications are:

1. Waiting Time Analysis:

➢ Service times in queuing systems (like customer service centers)
➢ Time between equipment failures in reliability engineering
➢ Duration of repair/maintenance activities

2. Financial Applications:
➢ Asset returns modeling in investment analysis
➢ Risk assessment in portfolio management
➢ Trading volume analysis in stock markets

1
 Module 4 – Probability Distributions | NA

➢ Duration of financial transactions

3. Medical and Biological Applications:

➢ Survival time analysis in medical studies
➢ Cell lifetimes in biological research
➢ Drug absorption rates in pharmacokinetics
➢ Patient recovery times

4. Environmental Sciences:
➢ Wind speed distributions
➢ Pollution concentration levels
➢ River discharge volumes
➢ Environmental contamination levels

5. Manufacturing and Quality Control:

➢ Production cycle times
➢ Machine processing times
➢ Product lifetime modeling
➢ Material strength testing

Note: The exponential distribution predicts the waiting time until the very first event occurs
while the gamma distribution predicts the waiting time until the 𝑘 𝑡ℎ event occurs.

Theorem: If ∀ 𝑎 > 0, 𝜆 > 0, and 𝑋~𝐺(𝑎, 𝜆), then prove that

𝜆
(i) 𝐸(𝑋) = 𝑎
𝜆
(ii) 𝑉𝑎𝑟(𝑋) = 𝑎2
𝜆
1
(iii) 𝑀𝑋 (𝑡) = ( ) ; if 𝑡 < 𝑎 [MGF of gamma distribution]
𝑡
1−
𝑎

Proof:

2
 Module 4 – Probability Distributions | NA

Characteristic function of gamma distribution:

Example 4.1 In a certain city, the daily consumption of electric power in millions of kilowatt
1
hours can be treated as a random variable X having gamma distribution with parameters (2 , 3).
If the power plant of the city has a daily capacity of 12 million kilowatt hours. What is the
probability that this power supply will be inadequate on any given day?

Solution:

3
 Module 4 – Probability Distributions | NA

Example 4.2 Consumer demand for milk in a certain locality, per month, is known to be a
general Gamma (Erlang) random variable. If the average demand is 𝑨 liters and the most likely
demand is 𝒃 liters (𝒃 < 𝑨), what is the variance of the demand?

4
 Module 4 – Probability Distributions | NA

Solution:

5
 Module 4 – Probability Distributions | NA

5. Exponential distribution

Exponential distribution: A continuous random variable X assuming non-negative values

is said to have an exponential distribution with parameter 𝜆 > 0, (i.e.,) 𝑋~𝐸(𝜆) if its
probability density function is given by 𝑓(𝑥) = 𝜆 𝑒 −𝜆𝑥 , 𝑥 ≥ 0.

Moment generating function of exponential distribution:

6
 Module 4 – Probability Distributions | NA

7
 Module 4 – Probability Distributions | NA

Example 5.1 The mileage which car owners get with a certain kind of radial tire is a random
variable having an exponential distribution with mean 40,000 km. Find the probabilities that
one of these tires will last (i) at least 20,000 km and (ii) at most 30,000 km.

Solution:

Example 5.2 If the time 𝑇 to failure of a component is exponentially distributed with parameter
𝜆 and if 𝑛 such components are installed, what is the probability that one-half or more of these
components are still functioning at the end of 𝑡 hours?

Solution:

8
 Module 4 – Probability Distributions | NA

Example 5.3 The time (in hours) required to repair a machine is exponentially distributed with
1
parameter 𝜆 = 2. (a) What is the probability that the repair time exceeds 2 ℎ? (b) What is the
conditional probability that a repair takes at least 10 ℎ given that its duration exceeds 9 ℎ?

Solution:

9
 Module 4 – Probability Distributions | NA

Example 5.4 The life length 𝑋 of an electronic component follows exponential distribution.
There are 2 processes by which the component may be manufactured. The expected life length
of the component is 100 ℎ if process I is used to manufacture, while it is 150 ℎ if process II is
used. The cost of manufacturing a single component by process I is Rs. 10, while it is Rs. 20
for process II. Moreover, if the component lasts less than the guaranteed life of 200 ℎ, a loss
of Rs. 50 is to be borne by the manufacturer. Which process is advantageous to the
manufacturer?

Solution:

10
 Module 4 – Probability Distributions | NA

Example 5.5 The length of telephone conversation is an exponential variate with mean 3 mins
and find the probability that call (a) end in less than 3 mins. (b) takes between 3 to 5 mins.

Solution:

Example 5.6 A random variable X has an exponential distribution with pdf

11
 Module 4 – Probability Distributions | NA

Solution:

Example 5.7 A memoryless property of the exponential distribution.

Statement:

Proof:

1
Example 5.8 The life of an electric bulb is exponentially distributed with failure rate 𝜆 = 3
(one failure in every 3000 hours on an average). Find 𝑃(𝑋 > 3.5|𝑋 > 2.5)?

Solution:

12
 Module 4 – Probability Distributions | NA

6. Weibull distribution

Weibull distribution: A continuous random variable X is said to follow a Weibull

distribution with parameter 𝛼, 𝛽 > 0, (i.e.,) 𝑋~𝐸(𝜆) if its probability density function is given
𝛽
by 𝑓(𝑥) = 𝛼 𝛽 𝑥 𝛽−1 𝑒 −𝛼 𝑥 , 𝑥 > 0, 𝛼, 𝛽 > 0.

Example 6.1 Each of the 6 tubes of a radio set has a life length (in years) which may be
considered as a RV that follows a Weibull distribution with parameters 𝛼 = 25 and 𝛽 = 2. If
these tubes function independently of one another, what is the probability that no tube will have
to be replaced during the first 2 months of service?

Solution:

13
 Module 4 – Probability Distributions | NA

Example 6.2 If the life 𝑋 (in years) of a certain type of car has a Weibull distribution with the
parameter 𝛽 = 2, find the value of the parameter 𝛼, given that probability that the life of the
car exceeds 5 years is 𝑒 −0.25. For these values of 𝛼 and 𝛽, find the mean and variance of 𝑋.

Solution:

Example 6.3 If the time 𝑇 to failure of a component follows a Weibull distribution with
parameters 𝛼 and 𝛽, find the hazard rate or conditional failure rate at time 𝑡 of the component.

Solution:

14
 Module 4 – Probability Distributions | NA

Example 6.4 The life time of a component measured in hours follows Weibull distribution with
parameters 𝛼 = 0.2, 𝛽 = 0.5. Find the mean life time of the component?

Solution:

Summary of the distributions:

Gamma Distribution:

• When to use: Modeling waiting times for multiple events, sum of exponential random
variables, or processes requiring multiple steps to complete.

• Origins: Derived from Euler's gamma function and developed to model waiting times
for multiple independent events.

• Example: "waiting for 𝑘 consecutive events to occur," such as waiting for 𝑘 customers
to be served in the system.

15
 Module 4 – Probability Distributions | NA

Exponential Distribution:

• When to use: Modeling time between random events in Poisson processes, systems
with constant failure rates, or memoryless processes.

• The memoryless property of exponential distributions means that "the future doesn't
depend on the past." In everyday terms:

Imagine waiting for a bus that arrives randomly. If buses follow an exponential
distribution and you've already waited 20 minutes, your chances of the bus arriving in
the next 5 minutes are exactly the same as if you had just arrived at the stop. The system
doesn't "remember" how long you've been waiting.

Other examples:
➢ A light bulb that's been working for 1000 hours has the same probability of failing in
the next hour as a brand-new bulb.

➢ In radioactive decay, an atom that has existed for years has the same chance of decaying
in the next second as a newly formed atom.

➢ This property is counterintuitive because most real things wear out or age. But for truly
random events (like radioactive decay or random customer arrivals), the exponential
distribution captures this "fresh start" property mathematically.

• Origins: Emerged from studying radioactive decay and random processes with no
memory.

• Examples: Demonstrating the memoryless property using waiting time (i.e., waiting
for buses).

Weibull Distribution:

• When to use: Modeling systems with varying failure rates, product lifetimes, or
extreme values.

• Origins: Developed by Waloddi Weibull (1951) initially for material strength

modeling.

• Examples: light bulb lifetimes, customer service times, or manufacturing quality

control, the shape parameter creates different failure patterns - early failures (infant
mortality), random failures, or wear-out failures.

16