CSE416 Applied Probability and Statistics

Module I: Probability and RVs

Topic: Axioms of Probability

Dr. Anil Chandra

 Theory of Probability
 Chance of unborn baby to be a boy or a girl?
 Chance of getting 1 in a throw of unbiased dice?
 Chance of a person getting hospitalized after catching
covid infection?
 Chance of NIFTY 50 index to reach 26,700 INR on
Thursday December 4, 2025? Note that on Friday,
November 28, 2025, it was 26,202.95 INR?
 Theory of Probability
 Random Experiments: Random Experiment is defined as an experiment in which:
• No personal bias is exercised while performing the experiment (Unbiased)
• The probabilities of occurrence of all possible outcomes can be determined
• We cannot predict in advance which outcome we will get in a trial
 Sample Space (Just like Universal Set U): In probability theory, the sample space of an
experiment is the set of all distinct possible outcomes or results of that experiment
• Eg, Rolling a dice = 1, 2, 3, 4, 5, 6
 Event and Trial: When a random experiment is performed under identical conditions, the
possible outcomes are known as Events and the experiment is called a trial.
 Mutually exclusive events: Two events are said to be mutually exclusive if there is no
element common in these two events or occurrence of one event excludes occurrence of
another event. Eg) Getting Queen and Getting Ace in a pack of 52 cards
 Independent events: Two events are said to be independent if occurrence of one event
does not affect occurrence of another event.
 Equally Likely Events: Two events are said to be equally likely if occurrence of one event
cannot occur in preference to other event and each event has the same chance of occurring
in a trial. For eg. When we are throwing a dice, getting 1 cannot occur in preference to
getting 2, getting 2 cannot occur in preference to getting 3 and so on;
 Axioms of Probability
 A.N. Kolmogorov, Russian Mathematician published his work Foundations of
Probability Theory (1934, German)(1956, Eng)
 Defined the theory of probability over continuous subsets of sample Space S or Ω
 Probability for any random variable x i.e. P(x) is continuous and P(A) is between
[0 1]
 It follows following three axioms:
S or Ω
(i) Non-Negativity: 0 ≤ P(A)
(ii) Normalization: P(S) = 1
(iii) Additivity: If A1, A2, A3, …, An are n
mutually exclusive events, then A1 A2 A3
P(A1UA2UA3U… UAn) =
P(A1)+P(A2)+ P(A3)+…+P(An)
 Importance of these Axioms of Probability
 These Axioms provided a Framework for theory of probability, which
overcame the limitations of classical and statistical approach to
probability. These limitations were that:
• The classical approach assumed “Sample Space” to be a finite space and all the
events in a sample space to be equally likely to occur
• Statistical approach which was based on statistical regularity of data, assumed
“Sample Space” to be a finite space
 These Axioms helped in further development of important theorems of
probability like Theorem of Total Probability and Baye’s Theorem
Practical example to show equally likely events

Shooter 1 0

2 1
2
3
4
5

Equally Likely says P(Each point) = 1/6

This is not true for a dart-throwing game
It is also not true for many real life cases
 Practical example to show dependency of events

Shooter 1 – 1st Trial Shooter 1 – 100th Trial

1 2 1
2 2
3 3
4 4
5 5

With each trial the probability of hitting target increases due to human learning curve over
time/trials
Hence probability of hitting 4 when 10th trial is done is dependent upon probability of 9th trial
But after, say 100000 trials, the probability of hitting 4 will become independent of previous
trial. But, Why?
 What we have learned from dart hitting game?

 Concept of Equally likely events, Independent and Dependent

Events, Biasedness in data
 Frequentist approach to determine probability
 Law of large numbers
 Discrete probability distribution
 Uniform distribution
 Continuous probability distribution
 Normal distribution
CSE416 Applied Probability and Statistics
Module I: Probability and RVs
Topic: Conditional probability

Dr. Anil Chandra

 Conditional Probability
 When events are dependent on each other  Concept
of Conditional Probability
 One example is evident from “dart game”
 The concept is used in Spam E-mail classification
• Use of certain words like “lottery”, “apply now”
• Contains links to malicious webpages
• Etc.
 Also used in Natural Language Processing etc.
• Dr. ABC is a good ….. (Doctor, Teacher, Orator ….)
 CONDITIONAL PROBABILITY
 When the events are dependent on each other, the concept of
conditional probability comes into picture
 Let S be a sample space of a random experiment. Let A and B be
any two events of S
 Then probability of occurrence of event A under the condition that
(i) event B has already occurred & (ii) P(B) ≠ 0 is given by:
P (A | B) = P (A∩B)
P(B)
 Similarly probability of occurrence of event B under the condition that
(i) event A has already occurred & (ii) P(A) ≠ 0 is given by:
P (B | A) = P (A∩B)
P(A)
 CONDITIONAL PROBABILITY - Application and Approach to
Baye’s Theorem – Detection of Spam e-mail
 Let S be a sample space of a random experiment – “App for Detection of Spam E-mail”
 Let event A be “probability of e-mail being a spam”
 Let event B be the word “lottery” in e-mail text
 Then P (A | B) indicates “probability that e-mail is a spam” when it is given that the
word “lottery” is mentioned in the e-mail, which may be determined as:
P (A | B) = P (A∩B) = Probability of e-mail being spam and containing text “lottery” ---(1)
P(B) Probability that e-mail contains the word “lottery” ---(2)
Condition is P(B)≠0 (i.e. it must contain the word “lottery”)
 Similarly, P (B | A) indicates probability that word “lottery” in mentioned in the e-mail
when it is given that the “e-mail is spam”, which may be determined as:
P (B | A) = P (A∩B) = Probability of e-mail being spam and containing text “lottery”
P(A) Probability that e-mail is spam
Condition is P(A)≠0 (i.e. must be receiving a spam email)
 Application of Baye’s Theorem – Spam e-mail detection

 Now let’s understand the Bayesian approach

 By Equations (1) and (2) of previous slides
P(A | B) = P(B | A) P(A)
P(B)
Hypothesis is P(A) = Probability that e-mail is spam (irrespective of whether it is detected by app as spam or not) – this
may contain false negatives (Hypothesis accepted as not spam when evidence suggests it is spam) i.e. e-mail is “spam”
but it may be detected by app as “not spam”
Evidence is P(B) = Probability that e-mail contains the word “lottery” (irrespective of whether it is spam or not) – this may
contain false positives (Hypothesis accepted as spam when evidence suggests it is not spam) i.e. e-mail may be detected
by app as “spam” but it is “not spam”
P(A | B) = Probability of “e-mail being spam” given that it contains the text “lottery” (probability of hypothesis “A”
being true given that evidence “B” is true) This probability REMOVES false positives
P(B | A) = Probability of e-mail containing the word “lottery” given that the “e-mail is spam”
(probability of evidence “B” being true given that hypothesis “A” is true). This probability REMOVES false
negatives
CSE416 Applied Probability and Statistics
Module I: Probability and RVs
Topic: Bayes’ theorem

Dr. Anil Chandra

 Baye’s Theorem – Basic Definition
 Thomas Baye’s (1701-1761), British Mathematician
 Suppose there are two events A and B which are dependent on each other. Then
P(A | B) = P (A ∩ B) / P(B) Hence P (A ∩ B) = P(A | B) P(B) … (1)
P(B | A) = P (A ∩ B) / P(A) Hence P (A ∩ B) = P(B | A) P(A) ….(2)
From equations (1) and (2), we get
P(A | B) P(B) = P(B | A) P(A) which implies
P(A | B) = P(B | A) P(A)
P(B)
P(A | B) = Posterior probability - the probability of hypothesis A being true, given that evidence B is true
P(B | A) = Likelihood - the probability of evidence B being true, given that hypothesis A is true
P(A) = Prior probability - the initial probability of hypothesis A being true before considering the evidence
P(B) = Marginal likelihood - the total probability of evidence B being true

This is also known as Posterior probability formula (Or Bayesian Approach) – used in
NLP Algorithms
 Example showing application of Baye’s Theorem

 Suppose a face recognition tool is being tested in a traffic stop to automatically detect if
a person driving a four-wheeler is wearing a seat-belt or not
 Let is name events as
• A  Person is not wearing seat-belt B  App has detected – Not wearing seat-belt
 Now let’s understand the Bayesian approach
P(A | B) = P(B | A) P(A)
P(B)
P(A | B) = Probability of person actually not wearing seat-belt (hypothesis A being true), when app has given a
positive result (given that evidence B is true) Removing false positive i.e. app detects that a person is not
wearing a seat-belt but he was wearing it
P(B | A) = Probability of app has given a positive result (evidence B being true) given that person is also not wearing
seat-belt (given that hypothesis A is true) Removing false negative i.e. person was not wearing seat-belt but he
was detected as having wearing it
P(A) = Probability that person is not wearing seat-belt (irrespective of whether app detects it or not) – this
may contain false negative i.e. person was not wearing seat-belt but he was detected as having wearing it
P(B) = Probability of app detecting that person is not wearing a seat-belt (irrespective of whether he is
wearing or not wearing a seat-belt) – this may contain false positives i.e. app detects that a person is not
wearing a seat-belt but he was wearing it
 Understanding Bayesian Approach - Calculations
(Source: Https://[Link]/chapter1/1_4_5_solved3.php)
 Box contains 4 coins out of which 3 are regular coins and 1 is fake coin with two heads
 A coin is picked at random and tossed
 What is the probability that Heads will show up? What if there
 It depends upon which coin is selected regular (say C1) and fake (say C2) are 2 coins
out of which
 Random experiment has two steps 1) Selecting a coin 2) Tossing it one is
 Probability that regular coin is selected ¾ fake coin is selected ¼ regular other
 If regular coin is selected, Probability of getting Heads = ½ is fake?
 If fake coin is selected, Probability of getting Heads = 1
 Probability of Getting Heads = (Regular Coin is selected & Tossed to get Heads) or (Fake coin
is selected & Tossed to get Heads)
P(H) = P(C1) P(H I C1) +P(C2) P(H I C2) = ¾ X ½ + ¼ X 1

Image source: [Link]

 Extension of Baye’s Theorem
 Thomas Baye’s (1701-1761), British Mathematician
 Box contains 4 coins out of which 3 are regular coins and 1 is fake coin with two heads
 A coin is picked at random and tossed
 Earlier question  What is the probability that Heads will show up?
 Revised question  If Heads show up what is the probability that it is a fake coin?
 P(C2 I H) = ?
 Favourable case = Fake Coin is Selected and Tossed to get Heads
 Total Cases = Heads Show Up
= Regular Coin is Selected and Tossed to get Heads OR Fake Coin is Selected and
Tossed to get Heads
Total Cases = P(C1) P(H I C1) +P(C2) P(H I C2)
 Favourable Case = P(C2) P(H I C2)
 P(C2 I H ) = Favourable Cases / Total Cases
= . P(C2) P(H I C2) .
P(C1) P(H I C1) +P(C2) P(H I C2)
 Baye’s Theorem for n mutually exclusive events

Note:
(1)“A” Has Occurred
(2) To find probability of any of Ei when A has occurred
 Question from previous year exams

Q 1) A fair dice is rolled twice. Define events,

A = “First roll is even”, B = “sum of outcomes is 7”
 Find P (A), P(B) and P (A | B)
 Verify Baye’s Theorem for these events (Internals/Mid Term Sept 2025)

Q 2) In a bolt factory, there are four machines A, B, C, D manufacturing 20%,

15%, 25% and 40% of the total output. Of their outputs 5%, 4%, 3% and
2%, in the same order, are defective bolts. A bolt is chosen at random from
thy factory’s production and is found defective. What is the probability that
the bolt was manufactured by machine A or machine D. (End Term Nov 2025)
𝑃(𝐸1) = 20% 𝑃(𝐸2) = 15%, 𝑃(𝐸3) = 25%, 𝑃(𝐸4) = 40%
Let A = Event of defective bolt,
𝑃(𝐴|𝐸1) = 0.05, 𝑃(𝐴|𝐸2) = 0.04 𝑃(𝐴|𝐸3) = 0.03 𝑃(𝐴|𝐸4) = 2% = 0.02
Total probability = 𝑃(𝐴) = 𝑃(𝐴|𝐸1)𝑃(𝐸1) + 𝑃(𝐴|𝐸2)𝑃(𝐸2) + 𝑃(𝐴|𝐸3)𝑃(𝐸3) + 𝑃(𝐴|𝐸4)𝑃(𝐸4) = 0.0315
𝑁𝑜𝑤, 𝑃(𝐸1 ∪ 𝐸4|𝐴) = 𝑃(𝐸1|𝐴) + 𝑃(𝐸4|𝐴) =
𝑃(𝐴|𝐸1)𝑃(𝐸1 ) + 𝑷(𝑨|𝑬𝟒)𝑷(𝑬𝟒) = 0.57
𝑷(𝑨) P(A)
CSE416 Applied Probability and Statistics
Module I: Probability and RVs
Topic: Random variables,
probability mass and density
functions,
expectation, variance
Probability Distribution

Dr. Anil Chandra

 Probability Distribution Definition
 Random Variable
• If X is a variable which takes the values {x1, x2, x3, ….., xn, ….}
with probabilities of their respective occurrences as {p1, p2, p3,
…., pn, …} & X is a part of a random experiment, then X is called
a random variable
 Probability Distribution
• A probability distribution is a statistical function that describes
all the possible outcomes that a random variable can take
within a given range and their respective probabilities of
occurrence. Or in other words, A probability distribution is a
table or an equation that links each outcome of a statistical
experiment with its probability of occurrence.
 TYPES PROBABILITY DISTRIBUTION

Discreet Continuous
Probability Distribution Probability Distribution
When Random Variable “X” When Random Variable “X”
is discrete is continuous

Meaning of discrete variable? Meaning of continuous variable?

• Variable which can take only finite number • Variables which can take all
of values values within any given
• Variables which are countable continuous interval
• Variables which can not take all values • Their values are infinite
within any given continuous interval but • Example) Height of individuals
rather can take some values i.e. has between 5’5” to 6”
interruptions or gaps in between
• Example) Printing mistakes in a book • P(X) = Probability Density
• P(X) = Probability Mass Function Function
 Discrete Probability Distribution
 Probability distribution wherein the variables are discrete random variables is called
Discrete Probability Distribution. Here P(X) is defined as Probability mass function
X (No. of Heads) 0 1 2 ….. ….. n
P(x) = Probability P(x=0) P(x=1) P(x=2) ….. …… P(x=n) TOTAL=1
of occurrence of X
Properties of a Discrete Probability Distribution
(i) 0 ≤ P(A) ≤ 1
(ii) ∑ P(x) = 1

Expected Value of X = E(X) = ∑ X.P(X) ….. (1)

E(X) is also referred to as Mean of the Probability Distribution

Variance of X = Var(X) = ∑ P(X).(X – E(X))2 which further simplifies to

Var (X)= E(X2) – [E(X)]2 , Here, E(X2)= ∑ (X2.P(X)) ….. (2)

Note Students are requested to remember equation number (1) and (2) for calculation of E(X) and Var(X) for
discrete probability distribution
 Application of two properties of discrete probability
distribution
 X is a discrete random variate having probability mass function:
x 0 1 2 3 4 5 6 7
P (X=x) 0 k 2k 2k 3k k2 2k2 7k2 +k
(i) Calculate k
(ii) Find P(X<6)
(i) Properties of a Discrete Probability Distribution
(i) 0 ≤ P(A) ≤ 1
(ii) ∑ P(x) = 1
∑ P(x) = 0+k+2k+2k+3k+k2+2k2+7k2+k = 10k2 + 9k (=1 by property ii)
So,10k2 + 9k = 1, if we solve this equation we get k = 1/10 OR k = -1
By property (i) k = 1/10
(ii) Now P(x < 6) = P(x=1) + P(x=2) + P(x=3) + P(x=4) + P(x=5)
= k2 + 8k = 8(1/10) + (1/10)2 = 0.81
NOTE  Also you can calculate Mean (Expected Value) & Variance
 Continuous Probability Distribution
 Probability distribution wherein the variables are continuous random variables is
called Continuous Probability Distribution.
 Variable x is continuous it can assume infinite number of values
 We can not draw any particular distribution table for a continuous probability
distribution.
 Due to this reason the probability mass function P(x) can not be defined in the
same manner as that we did in case of discrete probability distribution.
 In this case we introduce a function f(x) called probability density function
such that it satisfies the two conditions:
 Properties of Probability Density Function: If f(x) is a probability
density function of a continuous random variable X which is defined
on (-∞, +∞), then:
f (x) ≥ 0 for all x
+∞
∫f(x) dx = 1 (It means Area of Curve under whole distribution = 1)
 Expected value and Variance of Continuous
Probability Distribution
 If X is a continuous random variable with a Probability Density
Function (pdf) = f(x) defined on (-∞, +∞), then:

Let µ = Expected Value of X = …(3)

Variance of X =

Solving the above expression, we get

…(4)
Where …(5)

Note Students are requested to remember equation number (3) (4) and (5) for
calculation of E(X) and Var(X) for continuous probability distribution
 Example
 Literature Source: [Link]
_Probability_(Kuter)/4%3A_Continuous_Random_Variables/4.2%3A_Expected_Value_and_Variance_of_Continuous_Random_Variables

 Suppose X is a random variable which denotes waiting time of a person using an

elevator (in minutes) and pdf of X is given by:

Determine E(X) and V(X), use equations (3), (4) and (5) given in slide no. 27

Hence the expected time of a person in an elevator is 1 minute

Graphical representation of example

Expected Value of X
 Example (From Mid term Exams)
(a) Let X be a continuous RV with density function

(i) Find k
(ii) Compute E(X) and Var(X)
Hint: In this case, to find the value of k, use the property i.e.
Then use the following equations to find E(X) and Var(X)

(b) Define Probability mass Function (PMF) and Probability Density Function (PDF). List their
properties
CSE416 Applied Probability and Statistics
Module I: Probability and RVs
Topic: Binomial Distribution

Dr. Anil Chandra

 DISCRETE PROBABILITY DISTRIBUTION
- Binomial Distribution
• First defined by Swiss Mathematician Jacob Bernoulli (1654 - 1705)
• Parameters = n, p
• Where n = number of trials (each trials are independent)
• p = probability of success
• q = 1 – p is probability of failure (Either success or failure)

- Poisson Distribution
• First defined by French Mathematician SD Poisson (1781-1840)
• Same as above. Only exception:
• Parameter = λ (Average probability of occurrence of event)
• Where, λ = Expected value of random variable x
• Based on frequentist approach i.e. expected occurrences in the past
• No. of occurrences can be from 0 to infinity (theoretically)
• Talks about rare events (either p or q is very small)
 Examples – Binomial Distribution
 First defined by Swiss Mathematician Jacob Bernoulli (1654 - 1705)
 Parameters = n, p
 Where n = number of trials (each trials are independent)
 p = probability of success
 q = 1 – p is probability of failure (Either success or failure)
 For Example) Probability of passing an examination is 0.6. If 10 students appeared
in the examination. The consider the Random variable X:
• Random Variable X = No. of students appearing in the examination
• Finite or fixed number of trials = Finite (n=10)
• Each trial is independent of other = Yes
• There are only two possible outcomes (which are mutually exclusive, either success or
failure) = Yes
• Is it a rare event (either p or q is very small)= No, it is not a rare event since probability of
passing the examination is 60%
 Binomial Distribution B (n, p)
Let X be a discrete random variable. Let X ~ B(n, p) i.e. X follows a
binomial distribution. Then its probability mass function (PMF) is
defined as

P(x) = nCx px qn-x (x = 0,1,2,….,n)

where p & q are positive fractions called probability of occurrence of
event or probability of success and probability of non-occurrence of
that event or probability of failure respectively
n = number of trials
Parameters of the distribution = n, p
p+q=1
Mean = np = Expected Value of x
Variance = npq

First defined by Swiss Mathematician Jacob Bernoulli (1654 - 1705)

 Binomial Distribution Table values
P(x) = nCx px qn-x (x = 0,1,2,….,n)

P(x = 0) = nC0 p0 qn-0

P(x = 1) = nC1 p1 qn-1
P(x = 2) = nC2 p2 qn-2
...
...

P(x = (n-2)) = nCn-2 pn-2 q2

P(x = (n-1)) = nCn-1 pn-1 q1
P(x = n) = nCn pn q0
 Binomial Distribution – Example 1
Q) The overall percentage of failures in a certain exam is 40. What
is the probability that out of a group of 6 candidates, at least 4
passes the exam?
Sol) In this case it follows binomial distribution
p = probability of success = probability of passing in the exam = 0.6
q = probability of failure = probability of failing in the exam = 0.4
n=6
P (At least 4 passes the exam) = P(x≥4)
= P(x=4)+P(x=5)+P(x=6)
P(x=4) = 6C4 (0.6)4 (0.4)6-4
P(x=5) = 5C2 (0.6)5 (0.4)6-5
P(x=6) = 6C6 (0.6)6 (0.4)6-6
 Binomial – Example 2 from Mid term
A company manufactures safety helmets in large quantitites. From past records, it is
known that the probability of a helmet being defective is 0.1 (independent of others).
The quality-control department inspects a random sample of 20 helmets from a day’s
production.
(a) What is the probability that exactly 3 helmets are defective?
(b) What is the probability that at most 2 helmets are defective?
(c) Suppose the inspector decides to accept the lost if the sample contains no more
than 2 defectives. Find the probability that a lot is accepted. Interpret your result in
context of production quality

The probability that an AI tool named “BORE” is successful in a “bone-fracture

detection” by scanning X-Ray images is 0.9. Five images were scanned in “BORE”,
find the probability that it successfully detects bone-fracture in:
(a) No images
(b) At least three images
CSE416 Applied Probability and Statistics
Module I: Probability and RVs
Topic: Poisson Distribution

Dr. Anil Chandra

 Examples – Poisson Distribution
 Number of vehicles which malfunction in an express highway
 Finite or fixed number of trials = Theoretically infinite
 Each trial is independent of other = Yes
 There are only two possible outcomes (which are mutually exclusive,
either success or failure) = Yes
 Is it a rare event (either p or q is very small) = Yes (p = number of
vehicles which malfunction is very small as compared to number of
vehicles passing through the express highway)
 Expected value prediction = Expected Value of X
 Poisson Distribution P(λ)
 First defined by French Mathematician SD Poisson (1781-1840)
 Characterized by Parameter = λ (Average probability of occurrence of event)
 Based on frequentist approach i.e. expected occurrences in the past
 No. of occurrences can be from 0 to infinity (theoretically)
 Either p or q is very small as compared to the number of trails or population/sample size
 Let X be a discrete random variable and X ~ P(λ) i.e. X follows a Poisson
distribution with parameter λ (= Expected Value of the distribution) then its PMF is
defined as:
P(x) = e –λ λx
x!
where λ is the parameter of Poisson Distribution, λ ≥ 0
x = 0, 1, 2, 3, …. ∞
and e is an Euler’s number which is a constant = 2.718
Mean or Expected Value i.e. E(X) = λ
Variance i.e. V(x) = λ
 Poisson Distribution (continued)
 Sum of Poisson Variates: If X1, X2, …, Xk are k random variables and
Xi ~ P (λi) for all i = 1 to k, then X1 + X2 + … + Xk ~ P (λ1 + λ2 + …. + λk )
It means that sum of k number of Poisson variates is also a Poisson variate

 Theorem (Binomial approximation to Poisson process):

• If Yn~B (n, p), which means that Y is a Binomial Distribution with parameters n and p. Let λ
be a fixed real number, and limn∞np = λ,
then the PMF of Yn converges to a Poisson Distribution P(λ) as n∞ (large value of n) i.e.
lim PYn(x) = e –λ λx
n∞ x!
 Solved example
 Q1) Average number of accidents on any day on a national highway
is 1.8. Determine the probability that number of accidents are
• (i) At least one (ii) At most one [Given e-1.8 = 0.16529)
 (i) P(X ≥1) = P (X = 1) + P (X = 2) + P (X=3) + ….
 Since P(X ≥1) + P (X <1) = 1 (By Second Property of discrete
probability distribution)
 Hence P(X ≥1) = 1 – P (X <1) = 1 – P (X=0) = 1 - e-1.8 = 1 - 0.16529
 Example – Mid Term
A manufacturing company monitors customer complaints at its service center and the
weight variation of its packaged product. On average, the service center receives 3
complaints per hour (Poisson distributed). Find the probability that in a given 2-hour
period:
(a) No compliant is received
(b) At least 5 complaints are received
Hint  Calculate adjusted mean (λ) for 2-hour period

Suppose number of errors detected in a code is a random variable which follows a

Poisson distribution with average rate of 10 errors per 1000 lines of code.
(i) Find average rate of errors for 200 line-section of a code
(ii) Find the probability of having no errors in a 200-line section of a code
Hint  Calculate adjusted mean (λ) for 200 line-section of code
 Theorem: Binomial approximation to Poisson
If Yn~B (n, p), which means that Y is a Binomial Distribution with
parameters n and p.
Let λ be a fixed real number, and
limn∞np = λ
then the PMF of Yn converges to a Poisson Distribution P(λ) as
n∞ (large value of n) i.e.
lim PYn(x) = e –λ λx
n ∞ x!
 Eg. – Binomial Approximation to Poisson

Problem,: 5% of Deepavali fairy lights manufactured by a company are defective. The company's
Quality Control Manager is quite concerned and therefore randomly selects samples 100 bulbs
coming off of the assembly line. Let X denote the number in the sample that are defective. What is
the probability that the sample contains at most three defective bulbs?
Solving using Binomial Distribution: P(X≤3) = P(X=0)+P(X=1)+P(X=2)+P(X=3)
= 100C0 (0.05)0(0.95)100 + 100C1 (0.05)1(0.95)99+ 100C2 (0.05)20.95)98 + 100C3 (0.05)3(0.95)97

Let’s try Binomial Approximation to Poisson Theorem.

Here p is very small (0.05) as compared to n = 100 (which is very large)
Hence B(100, 0.05) ~ P(λ)
Since λ = E(X) = Expected Value of B(100, 0.05) = np = 100 X 0.05 = 5, i.e. λ=5
Hence B(100, 0.05) ~ P(5)
Hence, P(X≤3) = P(X=0)+P(X=1)+P(X=2)+P(X=3) =
e-5 50 + e-5 51 + e-5 52 + e-5 53
0! 1! 2! 3!
Since e-5 = 0.0067, the calculations will be easier in this case Answer = 0.265
You can crosscheck using online tools that Answer by using Binomial distribution will be 0.258 which is
close to the probability value by using Binomial Approximation to Poisson distribution
Note  The two conditions (i) p is very small and (ii) n is very large must be mut for this approximation to work
 Questions for practice
 The following table shows number of customers returning the
products in 100 stores. Fit a Poisson Distribution

No. of returns 0 1 2 3 4 5 6
No. of stores 4 14 23 23 18 9 9

 The normal rate of infection of a certain disease in animals is known

to be 25%. In an experiment with 6 animals injected with a new
vaccine it was observed that none of the animals caught infection.
Calculate probability of observed result. (Binomial Distribution)
CSE416 Applied Probability and Statistics
Module I: Probability and RVs
Topic: Geometric Distribution

Dr. Anil Chandra

 Geometric Distribution G(p)
 Geometric distribution is a discrete probability distribution that defines the number of
trials required to get the first success in a series of independent and identically
distributed Bernoulli trials, where each trial has two possible outcomes: success or
failure.
 Sometimes called the Furry distribution after physicist Wendell H. Furry.
 However, geometric series has roots in ancient Greek Mathematics
 Characterized by Parameter = p (probability of success in a single trial)
 Let X be a discrete random variable and X ~ G(λ) i.e. X follows a Geometric
distribution with parameter p then its PMF is defined as:
P(x) = (1 – p)(x – 1).p
where p is probability of success in a single trial
x = 0, 1, 2, 3, …. ∞
Mean or Expected Value i.e. E(X) = 1/p
Variance i.e. V(x) = 1 – p
p2
 Examples - Geometric Distribution
 Geometric distribution deals with finding the probability of needing a certain number
of attempts to achieve the first success
 It is sometimes referred to as a “waiting distribution”
 Like, the number of coin flips to get the first heads, the number of dice rolls to
get a specific number say number 6.
 Dart Game: Number of attempts to achieve throw of bulls-eye
 Patient waiting for a suitable kidney donor. Number of attempts to get a suitable
donor can be modeled through Geometric distribution
 Dating App: - Number of swipes or potential matches a person may go
through before a success (finding a mutual match) – It may be modeled
theoretically to estimate the number of trials to achieve the success.
Note  Geometric distribution has memoryless property , which means that probability of future
success is independent of past failures, i.e. probability of success (say, getting heads in a toss of
coin = 50%) remains same for the next trial even if the present trial is failure (conditional
probability)
Caution The real-world data it may not follow some of the required assumptions, like
probability of success is constant in each trial (swipe for match) (matching with a donor). Still
geometric distribution helps in estimating the expected values
 Geometric Distribution - Questions
 If a patient is waiting for a suitable blood donor and the probability that the
selected donor will be a match is 0.2, then find the expected number of
donors who will be tested till a match is found, including the matched donor.
• Given, p = 0.2; E[X] = 1 / p = 1 / 0.2 = 5
• The expected number of donors who will be tested till a match is found is 5
 Suppose you are playing a game of darts. The probability of success is 0.2.
What is the probability that you will hit the bullseye on the third try?
• Given, p = 0.2
• P(X = x) = (1 - p)x – 1.p
• P(X = 3) = (1 - 0.2)3 - 1(0.2) = (0.8)2(0.2) = 0.128
• The probability that you will hit the bullseye on the third try is 0.128
CSE416 Applied Probability and Statistics
Module I: Probability and RVs
Topic: Exponential Distribution

Dr. Anil Chandra

 Exponential Distribution E(λ)
 Exponential distribution is a continuous probability distribution. If X is a continuous
random variable which follows an Exponential Distribution i.e. X ~ E(λ), then the
probability density function (pdf) of X is defined as:

Where λ is defined as a rate parameter (how often the event occurs)

Mean or Expected Value i.e. E(X) = 1/λ
Variance i.e. V(x) = 1/λ2
 Exponential Distribution - Examples
 Lifespan of a product like an electronic appliance, say Refrigerator: The time until
the appliance fail, This can be modelled by exponential distribution
 Waiting time for a customer in a queue can be modelled by this distribution
 Time between customer calls at a business center can be modelled by this
distribution
 In general, Exponential Distribution is used for analyze life time data, predict the
amount of waiting time until next event (success, failure, arrival etc.)
Note  Just like Geometric distribution, Exponential Distribution also has
“memoryless” property i.e. the probability of future event is independent of the
time already passed.
A practical example is that if a light bulb has already lasted for 100 hours, the
probability of failing it in the next hour is same as the probability of new bulb
failing in its first hour (age does not affect its future performance). Hence it is a
conditional probability of given that the product has survived for 100 hours, then
what is the probability that it will fail in next hour. This is applicable when the
product is in its useful life period and not true when the product reaches the
“wear-out” phase of its life-span
 Exponential Distribution - Questions
The lifetime of a machine component follows an exponential distribution with mean 500
hours. Find the probability that a component lasts:
(a) more than 600 hours
(b) between 400 and 600 hours
Hint  As E(X) = 1/λ Hence λ = 1/E(X) = 1/500 = 0.002
So, f(x) = λe – λx = 0.002 e – 0.002 x
P (X>600) =

P (400<X<600) =

The lifetime of an electronic component follows exponential distribution with mean life
expectancy of 500 hours. Find the probability that a randomly selected component of
this type exceeds its expected lifetime.
Hint  As E(X) = 1/λ Hence λ = 1/E(X) = 1/500 = 0.002
So, f(x) = λe – λx = 0.002 e – 0.002 x
 Examples
Literature Source: [Link]
 Consider a scenario where the waiting time for a customer to complete a
transaction at a bank's ATM follows an exponential distribution with a scale
parameter (λ) of 5 minutes. We want to find the probability that a customer will take
less than 3 minutes to complete the transaction.

 Suppose the claims at a life insurance company's 24-hour call centre occur at a 4-
per-hour rate. What is the probability that the next call arrives after more than 2
hours?
CSE416 Applied Probability and Statistics
Module I: Probability and RVs
Topic: Gamma Distribution

Dr. Anil Chandra

 Gamma Function and its properties
For any positive real number α, Gamma Function is defined as:

Properties:

Hence, by Property (1) and (4)

and so on

Also, By property (3) and (5)

 What Gamma Function does?
 Suppose we have to calculate 3.2! or i!, then by making use of gamma function

 Hence, a Gamma function extends the factorial function (n!) from positive
integers to real and complex numbers
 Gamma distribution Γ(α, β)
 If X is a continuous random variable which follows an Gamma Distribution with shape parameter
α>0 and rate parameter β>0 i.e.
X ~ Γ(α, β),
 then the PDF of X is defined as:

Properties
E(X) = α/β
Var(x) = α/β2

Note 
(i) Transformation of Gamma Distribution to Exponential Distribution: If we let α = 1, then
“Gamma Distribution” transforms to “Exponential Distribution”
(ii) Approximation of Gamma Distribution to Normal Distribution: As the shape parameter (𝛼) and
rate parameter (β) become larger in a gamma distribution, the distribution becomes more oncentrated
around its mean, the mean and variance increase, and its shape approaches that of a normal
distribution (bell-shaped).
(iii) Approximation to Erlang distribution: If shape parameter is a positive integer then Gamma
distribution transforms to Erlang distribution
 Gamma Distribution Γ(α, β) Example
 Suppose we have to model the time until an event occurs, especially when the event is part of a
process that happens at a constant rate (Poisson process), such as waiting times, and the life
span of Mechanical parts. Then this can be modelled using a Gamma Distribution

 Gamma Distribution often involves waiting times for multiple events, like a call center getting
calls or a process completing tasks. using shape (α) and rate (β) parameters to find
probabilities, such as the chance of waiting less than 5 minutes for the 3rd call when calls
arrive at 2 per minute

 The probability of waiting for 10 minutes for my order in a restaurant, if I am fourth in line?
 Gamma Distribution - Examples
Source: [Link]
 A server receives requests at an average rate of 2 per minute (λ=2). We want
to find the probability that the time to receive the third request (α=3) is less
than 5 minutes (x=5).
CSE416 Applied Probability and Statistics
Module I: Probability and RVs
Topic: Uniform Distribution

Dr. Anil Chandra

 Discrete Uniform Distribution U(a, b)
Uniform distribution may be a discrete probability distribution or a continuous probability
distribution.
If X is a discrete RV , then X is said to be uniformly distributed if each one of its events are
equally likely to occur i.e. the PMF is defined as
P (X = xi) = 1/n for each i = 1,, 2, 3, …., n
E(X) = (a + b)/2, where a is the minimum value and b is the maximum value
Var(X) = (b – a)2/12
Example: Rolling fair dice or flipping fair coins,
where each outcome has an equal probability.
Consider rolling a fair dice.
Minimum value a = 1, maximum value b = 6.
E(X) = (1 + 6)/2 = 3.5
Var(X) = (6 – 1)2/12

(Image Source: [Link] )

 Continuous Uniform Distribution U(a, b)
Literature & Image Source: [Link]
If X is a continuous RV defined over (-∞, ∞) then X is said to be uniformly distributed or X~U(a, b),
if its PDF is defined as:

E(X) = (a + b)/2
Var (X) = (b – a)2/12

Example) Random number generator

(To generate any number between 0 –1)
 Uniform Distribution - Numerical
 Literature Source: [Link]
Q 1) A random variable X has a uniform distribution over (-2, 2),
(i) Find k for which P(X>k) = 1/2 (ii) Evaluate P(X<1) (iii) P[|X-1|<1]

Q 2) If X is uniformly distributed in (-1,, 4) then

(i) Its mean is ______________.
(ii) Its variance is ______________.

Q 3) Using the uniform distribution probability density function for random

variable X. in (0, 20), find P(3< X < 16).
 Here, a = 0, b =20
 f(x) = 1/(20 - 0) = 1/20
 P(3< X < 16) = (16 - 3) × (1/20) = 13/20

Q 4) A random variable X follows a uniform distribution over the interval [2,10].

Find the probability that X lies between 4 and 8.
CSE416 Applied Probability and Statistics
Module I: Probability and RVs
Topic: Normal Distribution

Dr. Anil Chandra

 Normal Distribution N (µ, σ2)
 Normal distribution or Gaussian distribution is a continuous
probability distribution. If X is a RV which follows a normal distribution
i.e. X ~N (µ, σ2), then its PDF is defined as:

where - ∞ ≤ x ≤ ∞
µ = Mean of the normal distribution
σ = Standard deviation of the normal distribution
Here
E(X) = µ
Var(X) = σ2
 Normal Distribution Curve
Mean = μ
Highest
Standard deviation = σ
Frequency

Area= 68.2%

34.1% 34.1%
13.6% 13.6%
95.4%

2.15% 99.7% 2.15%

μ
μ – 3σ μ – 2σ μ–σ μ+σ μ + 2σ μ + 3σ
 Normal Distribution – Example
Marks scored by students
Average Marks = 65
Standard deviation of marks = 10 Highest
Total Students = 100 Frequency

68.2%

34.1% 34.1%
13.6% 13.6%
95.4%

2.15% 99.7% 2.15%

65
65–30 65–20 65–10 65+10 65+20 65+ 30
 Normal Distribution
 Normal distribution presents a continuous bell shaped
normal distribution curve. It associates with continuous
variables such as height, weight, age etc.
 There is only one maximum peak
 The normal curve is symmetric over y axis
 The height of the normal distribution is maximum at its
Mean. Here Mean, Median and Mode coincides with each
other. Thus for normal distribution Mean = Mode = Median
 Area of total curve = 1
 Numerical Example
The height distribution of a group of 1000 men is normal with mean height 64.5” and
standard deviation 4.5”
(i) Find the number of men whose height is less than 69” but greater than 55.5”
(ii) Find number of men whose height is greater than 50??? (Conversion to Standard
Normal Distribution or determine Z Score)
Average height = 64.5”
Standard deviation of marks = 4.5”
Total Men = 1000

34.1% 34.1%
13.6% 13.6%

2.15% 2.15%

64.5
?? ?? 64.5–4.5 64.5+4.5 ?? ??
 Conversion of Normal Distribution N (µ, σ2)
to Standard Normal Distribution N(0, 1)
For the conversion, following transformation is applied
Z = (X – µ)
σ
This is referred to as Z-score

Here, Z ~ N(0,1) i.e. Z is a continuous RV with PDF defined as:

Here E(X) = 0
Var(X) = 1
 Standard Normal Distribution Curve N(0, 1)
Mean = 0
Highest
Standard deviation = 1
Frequency

Area= 68.2%

34.1% 34.1%
13.6% 13.6%
95.4%

2.15% 99.7% 2.15%

0
-3 -2 -1 1 2 3
 Additional Questions

Q1) If µ = 50; σ = 10 in a normal distribution, find

(i) P (50 ≤ X ≤ 80)
(ii) P (60 ≤ X ≤ 70)
(iii) P (30 ≤ X ≤ 40)

Z score is defined as:

Z = . X – Mean .
Standard Deviation
Z Table is given in the next slide
Z Score
 Convert Values in terms of Z Score
Q1) If µ (Mean) = 50; σ (Standard Deviation) = 10 in a normal distribution, find
(i) P (50 ≤ X ≤ 80)
(ii) P (60 ≤ X ≤ 70)
(iii) P (30 ≤ X ≤ 40)

Z score is defined as:

Z = . X – Mean .
Standard Deviation (SD)
(i) P (50 ≤ X ≤ 80) = P (50 - Mean ≤ X - Mean ≤ 80 - Mean) = P (50 - 50 ≤ X - Mean ≤ 80 - 50) = P(0 ≤ Z ≤ 3)
SD SD SD 10 SD 10
P(0 ≤ Z ≤ 3) = Half of 99.75% of the curve = 0.9975/2 = 0.49875
(ii) P (60 ≤ X ≤ 70) = P (60 - Mean ≤ X - Mean ≤ 70 - Mean) = P (60 - 50 ≤ X - Mean ≤ 70 - 50) = P(1 ≤ Z ≤ 2)
SD SD SD 10 SD 10
P(1 ≤ Z ≤ 2) = 13.6%= 0.136 (How ?? Refer to Z Table Or the Normal Distribution Curve)
 Numerical example from exams
A manufacturing company monitors customer complaints at its service center and the weight variation
of its packaged product. The weight of the product is normally distributed with mean = 500 gand
standard deviation = 10 g. The company sets an acceptable limit tolerance of ±15 g around the mean
(a) Find the probability that a randomly selected product will meet the weight specification
(b) If 1000 such products are produced in a day, how many would you expect to fall outside the
tolerance limits?
Given:
P (Z ≤ 0.0) = 0.5, P (Z≤0.5) = 0.6915, P (Z≤0.7) = 0.7580, P (Z≤1.0) = 0.8413,
P (Z≤1.2) = 0.8849, P (Z≤1.4) = 0.9192, P (Z≤1.5) = 0.9332, P (Z≤1.6) = 0.9452,
P (Z≤1.8) = 0.9641, P (Z≤2.0) = 0.9772

The average number of acres burned by forest and range fires in a large New Mexico county is 4,300 acres per year,
with a standard deviation of 750 acres. The distribution of the number of acres burned is normal. What is the
probability that between 2,500 and 4,200 acres will be burned in any given year?
 Numerical example from exams
 The average number of acres burned by forest and range fires in a large New Mexico
county is 4,300 acres per year, with a standard deviation of 750 acres. The distribution of
the number of acres burned is normal. What is the probability that between 2,500 and
4,200 acres will be burned in any given year?
 Solution:
 Mean 𝜇 = 4300, SD 𝜎 = 750.
 We want 𝑃(2500 ≤ 𝑋 ≤ 4200) for 𝑋 ∼ 𝑁(𝜇, 𝜎2)
Compute Z Score by using formula
Z = (X – µ)
σ
𝑧1 = 2500 − 4300 = -2.4
750
𝑧2 = 4200 − 4300 = -0.133
750
Using the table P(2500 ≤ 𝑋 ≤ 4200) = P(-2.4 ≤Z ≤-0.133) = 0.4401 or 44.01%
 Numerical for Module I

 Literature Source; Fundamentals of Mathematical Statistics (Second Edition) by SC Gupta

and VK Kapoor. Publisher Sultan Chand and Sons
Page 5.7, Example 5.3) Two dice are rolled. Let X denote the random variable which counts the
number of points on the upturned faces, construct a table giving non-zero values of the probability
mass function and draw the probability chart. Calculate E(X) and Var(X)
Hint: Make the sample space and you will see that, in this case the distribution table can be
computed as

X 2 3 4 5 6 7 8 9 10 11 12
P(X) 1/36 2/36 3/36 4/36 5/36 6/36 5/36 4/36 3/36 2/36 1/36

E(X) = ∑ X.P(X)
Var(X) = E(X2) – [E(X)]2
Where E(X2)= ∑ (X2.P(X))
 Example Literature Source;
Fundamentals of Mathematical Statistics (Second Edition) by SC Gupta and VK Kapoor. Publisher Sultan Chand &Sons

Page 5.8, Example 5.4) An experiment consists of three independent tosses of a fair coin. Let:
X = The number of heads, Y = The number of head runs, Z = The length of head runs, a head run being defined
as consecutive occurrence of at least two heads, its length then being the number of heads occurring together in
three tosses of the coin.
Find the probability function of (i) X, (ii) Y, (iii) Z, (iv) X+Y and (v) XY

Event HHH HHT HTH HTT THH THT TTH TTT

X 3 2 2 1 2 1 1 0
Y 1 1 0 0 1 0 0 0
Z 3 2 0 0 2 0 0 0
X+Y = U (say) 4 3 2 1 3 1 1 0
XY = V (say) 3 2 0 0 2 0 0 0

X 0 1 2 3 Y 0 1 Z 0 1 2 3
P(X) 1/8 3/8 3/8 1/8 P(Y) 5/8 3/8 P(Z) 5/8 0 2/8 1/8

U = X+Y 0 1 2 3 4 V=XY 0 1 2 3
P(U) 1/8 3/8 1/8 2/8 1/8 P(V) 5/8 0 2/8 1/8
 Numerical for Module I
Literature Source;
Fundamentals of Mathematical Statistics (Second Edition) by SC Gupta and VK Kapoor. Publisher Sultan Chand &Sons

Page 5.13, Example 5.5) The diameter of an electric cable, say X is assumed to be a continuous random variable
with pdf: f(x) = 6x(1 – x), 0 ≤x ≤ 1
(i) Check that f(x) is a p.d.f. (ii) Determine b such that P(X<b) = P(X>b) (ii) Find E(X) and Var(X)
Hint:

Page 5.13, Example 5.6) A continuous RV X has a pdf, f(x) = 3x2, 0 ≤ x ≤ 1

(i) Find a, if P (X≤a) = P(X>a), (ii) find b if P(X>b) = 0.05 (iii) Find E(X) and Var(X)

Page 5.15, Example 5.10) A continuous RV X has a pdf, f(x) = kx2 (1 – x3), 0 ≤ x ≤ 1 and k is a constant
(i) Find the value of k (iii) Using this value of k, find E(X) and Var(X)
 Numerical for Module I
Literature Source;
Fundamentals of Mathematical Statistics (Second Edition) by SC Gupta and VK Kapoor. Publisher Sultan Chand &Sons

Page 5.14, Eg. 5.7) Let X be a continuous RV with p.d.f.

(i) Find the value of a (ii) Using this value of a Find P(X≤1.5), E(X) and Var(X)

Page 5.18, Eg. 5.13) Find E(X) and Var(X) where X is a continuous RV with p.d.f.

Page 5.21, Eg. 5.19) Suppose that life in hours of a certain part of radio tube is a continuous RV X with PDF:

Find the probability that:

(i) All of three original tubes in a given radio set will have to be replaced during the first 150 hours of operation?
(ii) None of the three original tubes will have to be replaced in first 150 hours of operation
(iii) A tube will last for less than 200 hours if it is known that the tube is still functioning after 150 hours of service
 Numerical for Module I
Literature Source;
Fundamentals of Mathematical Statistics (Second Edition) by SC Gupta and VK Kapoor. Publisher Sultan Chand &Sons
 Numerical for Module I
Literature Source;
Fundamentals of Mathematical Statistics (Second Edition) by SC Gupta and VK Kapoor. Publisher Sultan Chand &Sons

Page 5.23, Eg. 5.21) The amount of bread (in hundreds of pounds) x, that a certain bakery is able to sell in a day is
found to be numerical valued random phenomenon, with PDF

(a) Find the value of k, such that f(x) is a PDF

(b) What is the probability that the number of pounds of bread that will be sold tomorrow is (i) more than 500 pounds,
(ii) less than 500 pounds (iii) between 250 and 750 pounds)
(c) Denoting by A, B and C, the events that the pounds of bread sold are in b(i), b(ii) and b(iii) respectively,
(i) find P (A | B), P (A | C)
(ii) Are “A” and “B” Independent events? Note  A and B are independent of each other if P(A∩B) = P(A).P(B)
(iii) Are “A” and “C” independent events?

Page 5.24, Eg. 5.22) The kms X in thousands of kms which car owners get with a certain type of tyre is a random
variable having PDF

Find the probabilities that one of these tyres will last (i) At most 10,000 kms (ii) between 16000 to 24000 kms
Find the expected life (in kms) of that tyre
 Numerical for Module I
Literature Source;
Fundamentals of Mathematical Statistics (Second Edition) by SC Gupta and VK Kapoor. Publisher Sultan Chand &Sons

 P. 8.5, Eg. 8.1) Ten coins are thrown simultaneously. Fiind the probability of getting at least
seven heads
 P. 8.5, Eg. 8.2) A and B play a game for which their chances of winning are in the ratio of
3:1. Find A’s chance of winning at least three games out of the five games played
 P. 8.6, Eg. 8.3) A cofee connoisseur claims that he can distinguish between a cup of
instant coffee and a cup of percolator coffee 7%% of the time. It is agreed that his clain will
be accepted if he correctly identifies at least 5 of the 6 cups. Find his chances of having
the claim (i) accepted (ii) rejected, when he does have to ability he claims
 P. 8.7, Eg. 8.6) A departmental store in a wrokds has 10 machines which nay need
adjustment from time to time during the day. Three of these machines are old, each
having a probability of 1/11 of needing adjustment during the day, and seven are new
having corresponding probabilities of 1/21. Assuming that no machine needs adjustment
twice on the same day, find the probabilities that on a particular day: (i) just 2 old and no
new machines need adjustment (ii) If just two machines need adjustment, they are of
same type
 Numerical for Module I
Literature Source;
Fundamentals of Mathematical Statistics (Second Edition) by SC Gupta and VK Kapoor. Publisher Sultan Chand &Sons

 P. 8.35, Eg. 8.32) A manufacturer of cotter pins knows that 5% of his product is defective. If
he sells cotter pins in boxes of 100 and guarantees that not more than 10 pins will be
defective, what is the approximate probability that a box will fail to meet the required
guaranteed quality?
 P. 8.36, Eg. 8.35) A manufacturer, who produces medicine bottles, find out that 0.1% of the
bottles are defective. The bottles are packed in boxes containing 500 bottles. A drug
manufacturer buys 100 boxes from the producer of bottles. Using Poisson distribution, find
how many boxes will contain: (i) No defective (ii) at least two defectives[Given e-0.5 = 0.6065)
 P. 8.38, Eg. 8.40 Suppose that the number of telephone calls coming into a telephone
exchange between 10 AM to 11 AM, say X is a random variable with Poisson distribution
with parameter 2. Similarly, the number of calls arriving between 11 AM and 12 Noon, say Y,
has a Poisson distribution with parameter 6. If X and Y are independent, what is the
probability that more than 5 calls come in between 10 AM to 12 noon?
 P. 8.38, Eg. 8.42) If X is a Poisson variate with parameter λ such that
P(X=2)=9P(X=4)+90P(X=6). Find (i) λ (ii) E(X)
 Fit a Poisson distribution to the following
X 0 1 2 3 4
f 192 100 24 3 1
 Numerical – Normal Distribution
 P. 9.15 Eg 9.3) X is normally distributed with mean of X is 12 and S.D. 4 (a) Find the probability
of the following: (i) P (X≥20) (ii) P(X≤20) (iii) P(0≤X ≤12) (iv) Find k if P(X>k) = 0.25
 P. 9.16 Eg 9.4) X is normally distributed with mean 30 and S.D. 5. Find the probability (i)
P(26≤X≤40) (ii) P(X≥45) and (iii) | X – 30 | > 5
 P. 9.17 Eg 9.5) The mean yield for one-acre plot is 662 kg with s.d. 32 kg. Assuming normal
distribution, how many one-acre plot in a batch of 1000 plots will have yield of (i) over 700 kg (ii)
below 650 kg
Literature Source: Probability and Statistics by Dr. Hari Arora (Katson Books)
 A sample of 100 dry battery cells tested to find the length of life produced the following results:
Mean = 12 hours, standard deviation = 3 hours
Assuming that the data is normally distributed, what percentage of battery cells are expected to
have life: (a) More than 15 hours (b) less than 6 hours © between 10 and 14 hours (Ans (a)
15.87 (b) 2.28 (c) 47.74)
 Assuming that the diameters of 1000 brass plugs taken consecutively from a machine form a
normal distribution with mean 0.7515 cm and standard deviation of 0.002 cm. Find the number of
plugs likely to be rejected if the approved diameter is 0.752 ± 0.004 cm (Ans: 52)
 Numerical – Bayes' Theorem
 Literature Source: Probability and Statistics by Dr. Hari Arora (Katson Books)
 Three urns contain 6 red, 4 black; 4 red, 6 black, and 5 red, 5 black balls respectively.
One of the urns is selected at random and a ball is drawn out of it. If the ball drawn is red
find the probability that it is drawn from the first urn (Ans: 2/5)
 An insurance company insured 2000 scooter drivers, 4000 car drivers and 6000 truck
drivers. The probability of an accident involving a scooter driver, car driver and a truck
driver is 0.01, 0.03 and 0.15 respectively. One of the insured person meets with an
accident. What is the probability that he is a scooter driver? (Ans: 1/52)
 A company has two plants to manufacture scooters, Plant I manufactures 70% of the
scooters and Plant II manufactures 30%. At Plant I, 80% of the scooters are rated as of
standard quality and at Plant II, 90% of the scooters are rated as of standard quality. A
scooter is chosen at random and is found to be of standard quality. What is the probability
that it has come from Plant II? (Ans: 27/83)
 By examining the chest X-ray, the probability that a person is diagnosed with TB when he
is actually suffering from it is 0.99. The probability that the doctor incorrectly diagnoses as
person to be having TB, on the basis of X-ray reports, is 0.001. In a certain city, 1 in 1000
persons suffers from TB. A person is selected at random and is diagnosed to have TB.
What is the chance that he actually has TB? (Ans: 110/221)
 References
References
 Fundamentals of Mathematical Statistics by SC Gupta and VK
Kapoor. Publisher Sultan Chand and Sons

89