Probability — 1

Amit K Biswas

Indian Statistical Institute

@ GLIM June-July 2021

For PGPM

July 27, 2024

Amit (ISI, Chennai) ProbDist July 27, 2024 1 / 47

 Introduction

Contents

1 Introduction
History
Random Events
Variables & Outcome
Measuring Probability

2 Axiom
Probability Types
Probability Rules

3 Conditional Probability
Independence
Bayes’ Law
Conditional Problems

Amit (ISI, Chennai) ProbDist July 27, 2024 2 / 47

 Introduction

Thus far . . .

So far we have discussed the following :

Types of variables : Discrete & Continuous, also subjective assigned score
like, taste of beer, score in beauty contest etc.
What is Data and it’s collection.
Compilation of data.
Measurement of centre and spread.
Population & sample.
Pictorial presentation — Histogram.
Skewness & Kurtosis.
Stratification & scales
We will now go on to discuss Probability.

Amit (ISI, Chennai) ProbDist July 27, 2024 3 / 47

 Introduction History

Contents

1 Introduction
History
Random Events
Variables & Outcome
Measuring Probability

2 Axiom
Probability Types
Probability Rules

3 Conditional Probability
Independence
Bayes’ Law
Conditional Problems

Amit (ISI, Chennai) ProbDist July 27, 2024 4 / 47

 Introduction History

Probability : The Subject

Jacob Bernouli (1654 − 1705), Abraham de Moivre (1667 − 1754), the

Reverend Thomas Bayes (1702 − 1761) and Joseph Lagrange
(1736 − 1813) developed probability formulae and techniques.

Amit (ISI, Chennai) ProbDist July 27, 2024 5 / 47

 Introduction History

Probability : The Subject

Jacob Bernouli (1654 − 1705), Abraham de Moivre (1667 − 1754), the

Reverend Thomas Bayes (1702 − 1761) and Joseph Lagrange
(1736 − 1813) developed probability formulae and techniques.

Probability was successfully applied at the gambling tables and more

relevantly to social and economic problems.

Amit (ISI, Chennai) ProbDist July 27, 2024 5 / 47

 Introduction History

Probability : The Subject

Jacob Bernouli (1654 − 1705), Abraham de Moivre (1667 − 1754), the

Reverend Thomas Bayes (1702 − 1761) and Joseph Lagrange
(1736 − 1813) developed probability formulae and techniques.

Probability was successfully applied at the gambling tables and more

relevantly to social and economic problems.

The mathematical theory of probability is the basis for statistical

applications in social, governmental, business and research
decision-making.

Amit (ISI, Chennai) ProbDist July 27, 2024 5 / 47

 Introduction History

Probability & Event

Probability is a way to quantify, understand and communicate

uncertainty. It measures the likelihood that something will occur or is
true (i.e., an event takes place).

Amit (ISI, Chennai) ProbDist July 27, 2024 6 / 47

 Introduction History

Probability & Event

Probability is a way to quantify, understand and communicate

uncertainty. It measures the likelihood that something will occur or is
true (i.e., an event takes place).

In Business or Management, we often study occurrence of events, for

example:

Amit (ISI, Chennai) ProbDist July 27, 2024 6 / 47

 Introduction History

Probability & Event

Probability is a way to quantify, understand and communicate

uncertainty. It measures the likelihood that something will occur or is
true (i.e., an event takes place).

In Business or Management, we often study occurrence of events, for

example:
1 how many more weeks before number of new cases fall below 10, 000 in
India?

Amit (ISI, Chennai) ProbDist July 27, 2024 6 / 47

 Introduction History

Probability & Event

Probability is a way to quantify, understand and communicate

uncertainty. It measures the likelihood that something will occur or is
true (i.e., an event takes place).

In Business or Management, we often study occurrence of events, for

example:
1 how many more weeks before number of new cases fall below 10, 000 in
India?
2 how many ventilators will be under breakdown/maintenance today?

Amit (ISI, Chennai) ProbDist July 27, 2024 6 / 47

 Introduction History

Probability & Event

Probability is a way to quantify, understand and communicate

uncertainty. It measures the likelihood that something will occur or is
true (i.e., an event takes place).

In Business or Management, we often study occurrence of events, for

example:
1 how many more weeks before number of new cases fall below 10, 000 in
India?
2 how many ventilators will be under breakdown/maintenance today?
3 how many iPhones will Apple sell in India in 2021 − 22?

Amit (ISI, Chennai) ProbDist July 27, 2024 6 / 47

 Introduction History

Probability & Event

Probability is a way to quantify, understand and communicate

uncertainty. It measures the likelihood that something will occur or is
true (i.e., an event takes place).

In Business or Management, we often study occurrence of events, for

example:
1 how many more weeks before number of new cases fall below 10, 000 in
India?
2 how many ventilators will be under breakdown/maintenance today?
3 how many iPhones will Apple sell in India in 2021 − 22?
4 what would be the GDP growth rate of India during the first quarter of
2021?

Amit (ISI, Chennai) ProbDist July 27, 2024 6 / 47

 Introduction History

Probability & Event

Probability is a way to quantify, understand and communicate

uncertainty. It measures the likelihood that something will occur or is
true (i.e., an event takes place).

In Business or Management, we often study occurrence of events, for

example:
1 how many more weeks before number of new cases fall below 10, 000 in
India?
2 how many ventilators will be under breakdown/maintenance today?
3 how many iPhones will Apple sell in India in 2021 − 22?
4 what would be the GDP growth rate of India during the first quarter of
2021?
5 Is the market likely to recover losses after the pandemic?

Amit (ISI, Chennai) ProbDist July 27, 2024 6 / 47

 Introduction Random Events

Contents

1 Introduction
History
Random Events
Variables & Outcome
Measuring Probability

2 Axiom
Probability Types
Probability Rules

3 Conditional Probability
Independence
Bayes’ Law
Conditional Problems

Amit (ISI, Chennai) ProbDist July 27, 2024 7 / 47

 Introduction Random Events

Randomness of phenomena

A phenomenon is random if individual outcomes are uncertain, but there is

nonetheless a regular distribution of outcomes in a large number of repetitions.

Amit (ISI, Chennai) ProbDist July 27, 2024 8 / 47

 Introduction Random Events

Randomness of phenomena

A phenomenon is random if individual outcomes are uncertain, but there is

nonetheless a regular distribution of outcomes in a large number of repetitions.

Examples :
1 Toss a fair coin, we get head or a tail.
2 Roll a die, any of the six faces may land up.
3 Waiting time at the vaccination centre : one may wait less than an hour
or longer.

Amit (ISI, Chennai) ProbDist July 27, 2024 8 / 47

 Introduction Random Events

Randomness of phenomena
There is still a regular pattern in the phenomenon, that is discovered only
after many repetitions.

Amit (ISI, Chennai) ProbDist July 27, 2024 9 / 47

 Introduction Random Events

Randomness of phenomena
There is still a regular pattern in the phenomenon, that is discovered only
after many repetitions.

Tossing a coin
The proportion of heads in ‘n‘ tosses of a coin changes as we keep tossing.
Eventually it approaches 0.5

Amit (ISI, Chennai) ProbDist July 27, 2024 9 / 47

 Introduction Random Events

Randomness of phenomena
There is still a regular pattern in the phenomenon, that is discovered only
after many repetitions.

Tossing a coin
The proportion of heads in ‘n‘ tosses of a coin changes as we keep tossing.
Eventually it approaches 0.5

Amit (ISI, Chennai) ProbDist July 27, 2024 9 / 47

 Introduction Random Events

Events & Random Variables

Consider the particulars of all employees in a company :

Serial ID Name Gender DOJ Age(yrs) Designation ..

1 00123 Selvi F 20-04-1989 55.5 Manager ..
2 00789 Ashokan M 07-05-1991 52.6 Officer ..
3 03331 Vinod M 15-07-1995 47.8 Cashier ..
4 03389 Chandra F 15-07-1995 49.1 Assistant ..
.. .. .. .. .. .. ..
. . . . . . . ..
5998 08776 Venkat M 05-02-2016 26.8 Clerk ..

Amit (ISI, Chennai) ProbDist July 27, 2024 10 / 47

 Introduction Variables & Outcome

Contents

1 Introduction
History
Random Events
Variables & Outcome
Measuring Probability

2 Axiom
Probability Types
Probability Rules

3 Conditional Probability
Independence
Bayes’ Law
Conditional Problems

Amit (ISI, Chennai) ProbDist July 27, 2024 11 / 47

 Introduction Variables & Outcome

Event & Random Variables

An interesting point is that while dealing with an event, we are concerned
with a random variable. In the above table, consider the heads/columns:
Employee ID, Name, Gender etc. All the values under any column
constitutes a sample space (S). Each such value is called an outcome. An
event (A) is a subset of S.

Amit (ISI, Chennai) ProbDist July 27, 2024 12 / 47

 Introduction Variables & Outcome

Event & Random Variables

An interesting point is that while dealing with an event, we are concerned
with a random variable. In the above table, consider the heads/columns:
Employee ID, Name, Gender etc. All the values under any column
constitutes a sample space (S). Each such value is called an outcome. An
event (A) is a subset of S.

In the ’Gender’ column, M & F are the two possibilities.

Amit (ISI, Chennai) ProbDist July 27, 2024 12 / 47

 Introduction Variables & Outcome

Event & Random Variables

An interesting point is that while dealing with an event, we are concerned
with a random variable. In the above table, consider the heads/columns:
Employee ID, Name, Gender etc. All the values under any column
constitutes a sample space (S). Each such value is called an outcome. An
event (A) is a subset of S.

In the ’Gender’ column, M & F are the two possibilities.

For example, if we are interested in the number of employees whose DOJ
is in 2001-2010, ’all the DOJs in 2001- 2010’ forms an event, and
’the number of employees whose DOJ is in 2010-2010’ is the concerned
random variable (typically denoted by capital letters: X , Y , Z etc.)

Amit (ISI, Chennai) ProbDist July 27, 2024 12 / 47

 Introduction Variables & Outcome

Event & Random Variables

An interesting point is that while dealing with an event, we are concerned
with a random variable. In the above table, consider the heads/columns:
Employee ID, Name, Gender etc. All the values under any column
constitutes a sample space (S). Each such value is called an outcome. An
event (A) is a subset of S.

In the ’Gender’ column, M & F are the two possibilities.

For example, if we are interested in the number of employees whose DOJ
is in 2001-2010, ’all the DOJs in 2001- 2010’ forms an event, and
’the number of employees whose DOJ is in 2010-2010’ is the concerned
random variable (typically denoted by capital letters: X , Y , Z etc.)

Again, an event is a description of the random variable over some range of

its outcomes. For example, {30 ≤ Z < 40} or {Z > 40} are events, where
Z denotes the age of an employees.

Amit (ISI, Chennai) ProbDist July 27, 2024 12 / 47

 Introduction Variables & Outcome

Event & Random Variables

An interesting point is that while dealing with an event, we are concerned
with a random variable. In the above table, consider the heads/columns:
Employee ID, Name, Gender etc. All the values under any column
constitutes a sample space (S). Each such value is called an outcome. An
event (A) is a subset of S.

In the ’Gender’ column, M & F are the two possibilities.

For example, if we are interested in the number of employees whose DOJ
is in 2001-2010, ’all the DOJs in 2001- 2010’ forms an event, and
’the number of employees whose DOJ is in 2010-2010’ is the concerned
random variable (typically denoted by capital letters: X , Y , Z etc.)

Again, an event is a description of the random variable over some range of

its outcomes. For example, {30 ≤ Z < 40} or {Z > 40} are events, where
Z denotes the age of an employees.

An important point is that outcomes in the sample space are generated

under similar condition.
Amit (ISI, Chennai) ProbDist July 27, 2024 12 / 47
 Introduction Variables & Outcome

Event — Examples

1 You are working for an automobile company. The company wants to

know how many times an automobile might fail during its warranty
period, say, before completing 10,000 KM of travel. The number of
failures is a random variable, say, X . An event {X ≤ 1} may be of
interest, i.e., at most one failure.)

Amit (ISI, Chennai) ProbDist July 27, 2024 13 / 47

 Introduction Variables & Outcome

Event — Examples

1 You are working for an automobile company. The company wants to

know how many times an automobile might fail during its warranty
period, say, before completing 10,000 KM of travel. The number of
failures is a random variable, say, X . An event {X ≤ 1} may be of
interest, i.e., at most one failure.)

2 Suppose an automobile has been serviced and you want to know how
many kilometers it would travel before encountering the next failure. If X
denote the number of kilometers of travel before failure, we are interested
in the event X ≥ 5000.

Amit (ISI, Chennai) ProbDist July 27, 2024 13 / 47

 Introduction Variables & Outcome

Event — Examples

1 You are working for an automobile company. The company wants to

know how many times an automobile might fail during its warranty
period, say, before completing 10,000 KM of travel. The number of
failures is a random variable, say, X . An event {X ≤ 1} may be of
interest, i.e., at most one failure.)

2 Suppose an automobile has been serviced and you want to know how
many kilometers it would travel before encountering the next failure. If X
denote the number of kilometers of travel before failure, we are interested
in the event X ≥ 5000.

In Management case studies or in Business Analytics, we always try to

identify the random variable and events of interest.

Amit (ISI, Chennai) ProbDist July 27, 2024 13 / 47

 Introduction Measuring Probability

Contents

1 Introduction
History
Random Events
Variables & Outcome
Measuring Probability

2 Axiom
Probability Types
Probability Rules

3 Conditional Probability
Independence
Bayes’ Law
Conditional Problems

Amit (ISI, Chennai) ProbDist July 27, 2024 14 / 47

 Introduction Measuring Probability

Probability of an Event

We use the symbol P(A) to denote the probability of an event A.

Let A be the event that a personal loan given to any customer turns out
to be bad (cannot be recovered). Suppose that the past record of the
bank shows that 3% of the personal loans turns out to be bad. Then we
write P(A) = 0.03, and interpret that under similar economic conditions,
a personal loan sanctioned today is likely to be bad with a probability of
0.03.

Amit (ISI, Chennai) ProbDist July 27, 2024 15 / 47

 Introduction Measuring Probability

Probability of an Event

We use the symbol P(A) to denote the probability of an event A.

Let A be the event that a personal loan given to any customer turns out
to be bad (cannot be recovered). Suppose that the past record of the
bank shows that 3% of the personal loans turns out to be bad. Then we
write P(A) = 0.03, and interpret that under similar economic conditions,
a personal loan sanctioned today is likely to be bad with a probability of
0.03.

Examples:
1 Suppose that 80% of small eateries go bust in a lockdown. This is
equivalent to saying that a randomly chosen eatery has a probability of
0.80 of going bust in case of a lockdown.

Amit (ISI, Chennai) ProbDist July 27, 2024 15 / 47

 Introduction Measuring Probability

Probability of an Event

We use the symbol P(A) to denote the probability of an event A.

Let A be the event that a personal loan given to any customer turns out
to be bad (cannot be recovered). Suppose that the past record of the
bank shows that 3% of the personal loans turns out to be bad. Then we
write P(A) = 0.03, and interpret that under similar economic conditions,
a personal loan sanctioned today is likely to be bad with a probability of
0.03.

Examples:
1 Suppose that 80% of small eateries go bust in a lockdown. This is
equivalent to saying that a randomly chosen eatery has a probability of
0.80 of going bust in case of a lockdown.
2 Over the past 100 trading days on the Bombay Stock Exchange, the
closing SENSEX index has fallen on 74 days, risen on 22 days and stayed
same on the remaining 4 days. Then the probability that it will fall on the
immediate next trading day is 74/100 = 0.74.

Amit (ISI, Chennai) ProbDist July 27, 2024 15 / 47

 Introduction Measuring Probability

Probability of an Event

Formalisation: If an experiment has n equally likely outcomes, r of which

produce the event A, then
Number of outcomes that produce A r
P[A] = =
Total number of outcomes n

Amit (ISI, Chennai) ProbDist July 27, 2024 16 / 47

 Introduction Measuring Probability

Probability of an Event

Formalisation: If an experiment has n equally likely outcomes, r of which

produce the event A, then
Number of outcomes that produce A r
P[A] = =
Total number of outcomes n

Exercise: Suppose that a telecom service provider carried out a survey

to find the level of importance its customer attach to various aspects of
their experience while using this service. A 7-point Likert scale is used for
rating, where 1 means least important and 7 stands for most important.
One aspect of the customer experience is accuracy in bills, and suppose
that the survey has yielded the following results.

Amit (ISI, Chennai) ProbDist July 27, 2024 16 / 47

 Introduction Measuring Probability

Probability of an Event – Example

Scale Value Frequency

1 1
2 3
3 6
4 13
5 72
6 135
7 130

Denote A to be the event that a randomly selected customer will consider the
importance of accuracy in billing to be 6 or more. What is the value of
P(A) ? How do you arrive at the value? What are the assumptions you made?

Amit (ISI, Chennai) ProbDist July 27, 2024 17 / 47

 Axiom

Contents

1 Introduction
History
Random Events
Variables & Outcome
Measuring Probability

2 Axiom
Probability Types
Probability Rules

3 Conditional Probability
Independence
Bayes’ Law
Conditional Problems

Amit (ISI, Chennai) ProbDist July 27, 2024 18 / 47

 Axiom

Axioms of Probability
A1. P(A ≥ 0) for any event A.

Amit (ISI, Chennai) ProbDist July 27, 2024 19 / 47

 Axiom

Axioms of Probability
A1. P(A ≥ 0) for any event A.

A2. P(S) = 1, where S is the sample space or the Universal set.

Amit (ISI, Chennai) ProbDist July 27, 2024 19 / 47

 Axiom

Axioms of Probability
A1. P(A ≥ 0) for any event A.

A2. P(S) = 1, where S is the sample space or the Universal set.

A3. If A ∩ B = Φ, (i.e., the events A and B are mutually exclusive), then

P(A ∪ B) = P(A) + P(B)

Amit (ISI, Chennai) ProbDist July 27, 2024 19 / 47

 Axiom

Axioms of Probability
A1. P(A ≥ 0) for any event A.

A2. P(S) = 1, where S is the sample space or the Universal set.

A3. If A ∩ B = Φ, (i.e., the events A and B are mutually exclusive), then

P(A ∪ B) = P(A) + P(B)

Amit (ISI, Chennai) ProbDist July 27, 2024 19 / 47

 Axiom

Axioms of Probability
A1. P(A ≥ 0) for any event A.

A2. P(S) = 1, where S is the sample space or the Universal set.

A3. If A ∩ B = Φ, (i.e., the events A and B are mutually exclusive), then

P(A ∪ B) = P(A) + P(B)

Important Properties:

P[Ac ] = 1 − P[A]

A ⊆ B ⇒ P[A] ≤ P[B]

Amit (ISI, Chennai) ProbDist July 27, 2024 19 / 47

 Axiom

Axioms of Probability
A1. P(A ≥ 0) for any event A.

A2. P(S) = 1, where S is the sample space or the Universal set.

A3. If A ∩ B = Φ, (i.e., the events A and B are mutually exclusive), then

P(A ∪ B) = P(A) + P(B)

Important Properties:

P[Ac ] = 1 − P[A]

A ⊆ B ⇒ P[A] ≤ P[B]

P[A ∪ B] = P[A] + P[B] − P[A ∩ B] when, A ∩ B ̸= Φ

Events A and B are independent if P[A ∩ B] = P[A].P[B]

Amit (ISI, Chennai) ProbDist July 27, 2024 19 / 47

 Axiom Probability Types

Contents

1 Introduction
History
Random Events
Variables & Outcome
Measuring Probability

2 Axiom
Probability Types
Probability Rules

3 Conditional Probability
Independence
Bayes’ Law
Conditional Problems

Amit (ISI, Chennai) ProbDist July 27, 2024 20 / 47

 Axiom Probability Types

Types of Probability

There are three basic ways of classifying probability. These three represent
rather different conceptual approaches to the study of probability theory.

Amit (ISI, Chennai) ProbDist July 27, 2024 21 / 47

 Axiom Probability Types

Types of Probability

There are three basic ways of classifying probability. These three represent
rather different conceptual approaches to the study of probability theory.

We have already talked about the first two.

Amit (ISI, Chennai) ProbDist July 27, 2024 21 / 47

 Axiom Probability Types

Classical Approach
Defines probability as ratio of favorable outcomes to the total outcomes.

Amit (ISI, Chennai) ProbDist July 27, 2024 22 / 47

 Axiom Probability Types

Classical Approach
Defines probability as ratio of favorable outcomes to the total outcomes.
Also known as priori probability.
It has a number of assumptions, hence is the most restrictive approach,
and it is least useful in real-life situations.

Amit (ISI, Chennai) ProbDist July 27, 2024 22 / 47

 Axiom Probability Types

Classical Approach
Defines probability as ratio of favorable outcomes to the total outcomes.
Also known as priori probability.
It has a number of assumptions, hence is the most restrictive approach,
and it is least useful in real-life situations.
Relative Frequency Approach
Defines probability as observed relative frequency of an event in a very
large number of trials.
It assumes less assumptions but requires the event to be capable of being
repeated a large number of times.

Amit (ISI, Chennai) ProbDist July 27, 2024 22 / 47

 Axiom Probability Types

Classical Approach
Defines probability as ratio of favorable outcomes to the total outcomes.
Also known as priori probability.
It has a number of assumptions, hence is the most restrictive approach,
and it is least useful in real-life situations.
Relative Frequency Approach
Defines probability as observed relative frequency of an event in a very
large number of trials.
It assumes less assumptions but requires the event to be capable of being
repeated a large number of times.
Subjective Probability
Deals with specific or unique situations typical of the business or
management world.
Based upon some belief or educated guess of the decision maker.
Subjective assessments of probability permits the widest flexibility of the
three concepts, also known as personal probability.

Amit (ISI, Chennai) ProbDist July 27, 2024 22 / 47

 Axiom Probability Rules

Contents

1 Introduction
History
Random Events
Variables & Outcome
Measuring Probability

2 Axiom
Probability Types
Probability Rules

3 Conditional Probability
Independence
Bayes’ Law
Conditional Problems

Amit (ISI, Chennai) ProbDist July 27, 2024 23 / 47

 Axiom Probability Rules

Probability Examples

Which one is more likely? Obtaining (a) at least one-SIX in 4 tosses of a fair
die, (b) at least one double-SIX in 24 tosses of a pair of dice.

Amit (ISI, Chennai) ProbDist July 27, 2024 24 / 47

 Axiom Probability Rules

Probability Examples

Which one is more likely? Obtaining (a) at least one-SIX in 4 tosses of a fair
die, (b) at least one double-SIX in 24 tosses of a pair of dice.
Answer: Sample space S =?

Amit (ISI, Chennai) ProbDist July 27, 2024 24 / 47

 Axiom Probability Rules

Probability Examples

Which one is more likely? Obtaining (a) at least one-SIX in 4 tosses of a fair
die, (b) at least one double-SIX in 24 tosses of a pair of dice.
Answer: Sample space S =?

Denote the random variables:

X = no. of SIX in 4 tosses of single die, and
Y = no. of double-SIX in 24 tosses of a pair of dice.

Amit (ISI, Chennai) ProbDist July 27, 2024 24 / 47

 Axiom Probability Rules

Probability Examples

Which one is more likely? Obtaining (a) at least one-SIX in 4 tosses of a fair
die, (b) at least one double-SIX in 24 tosses of a pair of dice.
Answer: Sample space S =?

Denote the random variables:

X = no. of SIX in 4 tosses of single die, and
Y = no. of double-SIX in 24 tosses of a pair of dice.

Define events A = {X ≥ 1} and B = {Y ≥ 1} . Then

Amit (ISI, Chennai) ProbDist July 27, 2024 24 / 47

 Axiom Probability Rules

Probability Examples

Which one is more likely? Obtaining (a) at least one-SIX in 4 tosses of a fair
die, (b) at least one double-SIX in 24 tosses of a pair of dice.
Answer: Sample space S =?

Denote the random variables:

X = no. of SIX in 4 tosses of single die, and
Y = no. of double-SIX in 24 tosses of a pair of dice.

Define events A = {X ≥ 1} and B = {Y ≥ 1} . Then

Ac = {X = 0}, similarly B c = {Y = 0}

Amit (ISI, Chennai) ProbDist July 27, 2024 24 / 47

 Axiom Probability Rules

Probability Examples

Which one is more likely? Obtaining (a) at least one-SIX in 4 tosses of a fair
die, (b) at least one double-SIX in 24 tosses of a pair of dice.
Answer: Sample space S =?

Denote the random variables:

X = no. of SIX in 4 tosses of single die, and
Y = no. of double-SIX in 24 tosses of a pair of dice.

Define events A = {X ≥ 1} and B = {Y ≥ 1} . Then

Ac = {X = 0}, similarly B c = {Y = 0}
54
P(A) = 1 − P(Ā) = 1 − 64 ≈ 0.518.
3524
P(B) = 1 − P(B̄) = 1 − 3624
≈ 0.0.491.
Question 1 to Poll

Amit (ISI, Chennai) ProbDist July 27, 2024 24 / 47

 Axiom Probability Rules

Probability Examples

1 Suppose three positions are open in a company for which 8 men and 7
women have applied for. Assuming each applicant is equally qualified for
either position, what is the probability that three men are selected for the
positions?

2 A problem in Statistics is given to three students A, B and C, whose

chances of solving it independently are 12 , 31 and 14 respectively. Find the
probability that (a) the problem is solved, (b) at least two of them solve
the problem.

Amit (ISI, Chennai) ProbDist July 27, 2024 25 / 47

 Conditional Probability

Contents

1 Introduction
History
Random Events
Variables & Outcome
Measuring Probability

2 Axiom
Probability Types
Probability Rules

3 Conditional Probability
Independence
Bayes’ Law
Conditional Problems

Amit (ISI, Chennai) ProbDist July 27, 2024 26 / 47

 Conditional Probability

Conditional Probability & Examples

Conditional probability of A given B is defined by

P[A∩B]
P[A|B] = P(B) , where P(B) > 0

Amit (ISI, Chennai) ProbDist July 27, 2024 27 / 47

 Conditional Probability

Conditional Probability & Examples

Conditional probability of A given B is defined by

P[A∩B]
P[A|B] = P(B) , where P(B) > 0

Some facts can be easily derived from the above definition:

If A and B are independent, then P[A|B] = P(A)
P[A|B] ̸= P[B|A]
P[Ā|B] = 1 − P[A|B]
P(A) = P[A ∩ B] + P[A ∩ B̄]

Question :

A six faced die has been tossed, I want you to tell me, the probability that an
even number has shown up. Your answer would be half. However, in addition
if I tell you, I have obtained a number more than 3, then??

Amit (ISI, Chennai) ProbDist July 27, 2024 27 / 47

 Conditional Probability

Conditional Probability & Examples

We know that probability of a heart attack during a period of say one year for
a randomly selected Indian male may be fairly low. However, this risk may
increase significantly for a given combination of age, genetic disposition,
smoking habit, BMI, level of blood sugar and LDL etc.

Amit (ISI, Chennai) ProbDist July 27, 2024 28 / 47

 Conditional Probability

Conditional Probability & Examples

We know that probability of a heart attack during a period of say one year for
a randomly selected Indian male may be fairly low. However, this risk may
increase significantly for a given combination of age, genetic disposition,
smoking habit, BMI, level of blood sugar and LDL etc.

Of the students in a certain college, 50% regularly attend the football games,
30% are first-year students and 40% are upper-class students who donate
regularly and attend football games.
1 What is the (conditional) probability that the person chosen, attends
football games regularly, given that s/he is first-year student?
2 What is the (conditional) probability that the chosen person is a first-year
student, given that s/he attends football games regularly?
Question 2 to Poll

Amit (ISI, Chennai) ProbDist July 27, 2024 28 / 47

 Conditional Probability

Conditional Probability & Examples

Suppose that eye colour of people in a group of 40 is as follows.

Black Green Total
Male 20 8 28
Female 10 2 12
Total 30 10 40

What is the probability that a randomly selected male has green eyes?

Amit (ISI, Chennai) ProbDist July 27, 2024 29 / 47

 Conditional Probability

Conditional Probability & Examples

Suppose that eye colour of people in a group of 40 is as follows.

Black Green Total
Male 20 8 28
Female 10 2 12
Total 30 10 40

What is the probability that a randomly selected male has green eyes?

Solution : Let M : Male, G : Green eyes.

To find : P[G |M] So using the definition :
∩M
P[G |M] = P[G 8/40
P[M] = 28/40

Amit (ISI, Chennai) ProbDist July 27, 2024 29 / 47

 Conditional Probability

Sensitivity & Specificity

Sensitivity is the probability that an individual with the COVID infection
(D) tests positive in RT-PCR,
Sensitivity = P[+ve|D].

Amit (ISI, Chennai) ProbDist July 27, 2024 30 / 47

 Conditional Probability

Sensitivity & Specificity

Sensitivity is the probability that an individual with the COVID infection
(D) tests positive in RT-PCR,
Sensitivity = P[+ve|D].

Specificity is the probability that an individual without COVID infection

(D c ) tests negative in RT-PCR,
Specificity = P[−ve|D c ].

Amit (ISI, Chennai) ProbDist July 27, 2024 30 / 47

 Conditional Probability

Sensitivity & Specificity

Sensitivity is the probability that an individual with the COVID infection
(D) tests positive in RT-PCR,
Sensitivity = P[+ve|D].

Specificity is the probability that an individual without COVID infection

(D c ) tests negative in RT-PCR,
Specificity = P[−ve|D c ].

Positive Predictive Value[PPV] = Probability that an individual having

tested positive has the disease D.
PPV = P[D| + ve]

Amit (ISI, Chennai) ProbDist July 27, 2024 30 / 47

 Conditional Probability

Sensitivity & Specificity

Sensitivity is the probability that an individual with the COVID infection
(D) tests positive in RT-PCR,
Sensitivity = P[+ve|D].

Specificity is the probability that an individual without COVID infection

(D c ) tests negative in RT-PCR,
Specificity = P[−ve|D c ].

Positive Predictive Value[PPV] = Probability that an individual having

tested positive has the disease D.
PPV = P[D| + ve]

Negative Predictive Value[NPV] = Probability that an individual having

tested negative does’t have the disease D.
PPV = P[−ve|D c ]
Amit (ISI, Chennai) ProbDist July 27, 2024 30 / 47
 Conditional Probability

Sensitivity & Specificity – Examples

A doctor advises a patient an RT-PCR test for COVID. Before the result of
the test, the only evidence the doctor can rely on is that 1 in 1000 patients
has COVID. Experience has shown that 99% of the cases in which COVID is
present, the test results in positive, and 95% of the cases in which it is absent,
the test result is negative. If the test turns out to be positive, what
probability the doctor would assign to the event that COVID is present?

Amit (ISI, Chennai) ProbDist July 27, 2024 31 / 47

 Conditional Probability

Sensitivity & Specificity – Examples

A doctor advises a patient an RT-PCR test for COVID. Before the result of
the test, the only evidence the doctor can rely on is that 1 in 1000 patients
has COVID. Experience has shown that 99% of the cases in which COVID is
present, the test results in positive, and 95% of the cases in which it is absent,
the test result is negative. If the test turns out to be positive, what
probability the doctor would assign to the event that COVID is present?
1 99 95
Answer : Given that P[D] = 1000 , P[+ve|D] = 100 and P[−ve|D c ] = 100 .

Amit (ISI, Chennai) ProbDist July 27, 2024 31 / 47

 Conditional Probability

Sensitivity & Specificity – Examples

A doctor advises a patient an RT-PCR test for COVID. Before the result of
the test, the only evidence the doctor can rely on is that 1 in 1000 patients
has COVID. Experience has shown that 99% of the cases in which COVID is
present, the test results in positive, and 95% of the cases in which it is absent,
the test result is negative. If the test turns out to be positive, what
probability the doctor would assign to the event that COVID is present?
1 99 95
Answer : Given that P[D] = 1000 , P[+ve|D] = 100 and P[−ve|D c ] = 100 .

P[+ve|D]P[D] 99
P[D| + ve] = c c
= = 0.0194
P[+ve|D]P[D] + P[+ve|D ]P[D ] 5094

Amit (ISI, Chennai) ProbDist July 27, 2024 31 / 47

 Conditional Probability Independence

Contents

1 Introduction
History
Random Events
Variables & Outcome
Measuring Probability

2 Axiom
Probability Types
Probability Rules

3 Conditional Probability
Independence
Bayes’ Law
Conditional Problems

Amit (ISI, Chennai) ProbDist July 27, 2024 32 / 47

 Conditional Probability Independence

Independence of events

Occurrence of one event has no effect on the probability of the occurrence

of other event.
There are three types of probabilities under statistical independence
Marginal — Probability of occurrence of the event, like P(A) or P(B)

Amit (ISI, Chennai) ProbDist July 27, 2024 33 / 47

 Conditional Probability Independence

Independence of events

Occurrence of one event has no effect on the probability of the occurrence

of other event.
There are three types of probabilities under statistical independence
Marginal — Probability of occurrence of the event, like P(A) or P(B)
Joint — A ∩ B : P(A ∩ B) = P(A) × P(B)

Amit (ISI, Chennai) ProbDist July 27, 2024 33 / 47

 Conditional Probability Independence

Independence of events

Occurrence of one event has no effect on the probability of the occurrence

of other event.
There are three types of probabilities under statistical independence
Marginal — Probability of occurrence of the event, like P(A) or P(B)
Joint — A ∩ B : P(A ∩ B) = P(A) × P(B)
Conditional — P(A|B) = P(A) and P(B|A) = P(B)
In statistical independence, assumption is that events are not related.

Amit (ISI, Chennai) ProbDist July 27, 2024 33 / 47

 Conditional Probability Independence

Independence of events

Occurrence of one event has no effect on the probability of the occurrence

of other event.
There are three types of probabilities under statistical independence
Marginal — Probability of occurrence of the event, like P(A) or P(B)
Joint — A ∩ B : P(A ∩ B) = P(A) × P(B)
Conditional — P(A|B) = P(A) and P(B|A) = P(B)
In statistical independence, assumption is that events are not related.
If we consider three events A, B and C , they are called pair-wise
independent if
P(A ∩ B) = P(A) × P(B)
P(B ∩ C ) = P(B) × P(C )
P(C ∩ A) = P(C ) × P(A)

Amit (ISI, Chennai) ProbDist July 27, 2024 33 / 47

 Conditional Probability Independence

Independence of events

Occurrence of one event has no effect on the probability of the occurrence

of other event.
There are three types of probabilities under statistical independence
Marginal — Probability of occurrence of the event, like P(A) or P(B)
Joint — A ∩ B : P(A ∩ B) = P(A) × P(B)
Conditional — P(A|B) = P(A) and P(B|A) = P(B)
In statistical independence, assumption is that events are not related.
If we consider three events A, B and C , they are called pair-wise
independent if
P(A ∩ B) = P(A) × P(B)
P(B ∩ C ) = P(B) × P(C )
P(C ∩ A) = P(C ) × P(A)
They are called mutually independent if, in addition
P(A ∩ B ∩ C ) = P(A) × P(B) × P(C )
Question 3 to poll

Amit (ISI, Chennai) ProbDist July 27, 2024 33 / 47

 Conditional Probability Independence

Independence and Exclusive

The difference between mutually exclusive and independent events is: a

mutually exclusive event can simply be defined as a situation when two
events cannot occur at same time whereas independent event occurs
when one event remains unaffected by the occurrence of the other
event.

Amit (ISI, Chennai) ProbDist July 27, 2024 34 / 47

 Conditional Probability Independence

Independence and Exclusive

The difference between mutually exclusive and independent events is: a

mutually exclusive event can simply be defined as a situation when two
events cannot occur at same time whereas independent event occurs
when one event remains unaffected by the occurrence of the other
event.

An example of a mutually exclusive event is when a single coin is a tossed

and there are two events that can occur, either it will be a head or a tail.
Hence, both the events here are mutually exclusive. But if we take two
separate coins and flip them, then the occurrence of Head or Tail on
both the coins are independent to each other.

Amit (ISI, Chennai) ProbDist July 27, 2024 34 / 47

 Conditional Probability Independence

Independence and Exclusive

Difference between Mutually exclusive and independent events.

Mutually exclusive events Independent events

1 When the occurrence is not When the occurrence of one
simultaneous for two events event does not control the
then they are termed as occurrence of the other event,
Mutually exclusive events. they are independent.

Amit (ISI, Chennai) ProbDist July 27, 2024 35 / 47

 Conditional Probability Independence

Independence and Exclusive

Difference between Mutually exclusive and independent events.

Mutually exclusive events Independent events

1 When the occurrence is not When the occurrence of one
simultaneous for two events event does not control the
then they are termed as occurrence of the other event,
Mutually exclusive events. they are independent.
2 The non-occurrence of an There is no influence of an
event will end up in the occurrence with another and .
occurrence of an event they are independent of each

Amit (ISI, Chennai) ProbDist July 27, 2024 35 / 47

 Conditional Probability Independence

Independence and Exclusive

Difference between Mutually exclusive and independent events.

Mutually exclusive events Independent events

1 When the occurrence is not When the occurrence of one
simultaneous for two events event does not control the
then they are termed as occurrence of the other event,
Mutually exclusive events. they are independent.
2 The non-occurrence of an There is no influence of an
event will end up in the occurrence with another and .
occurrence of an event they are independent of each
3 The mathematical formula for other. The mathematical formula for
mutually exclusive events can be independent events can be
represented as P(A ∩ B) = 0 defined as P(A ∩ B) = P(A).P(B)

Amit (ISI, Chennai) ProbDist July 27, 2024 35 / 47

 Conditional Probability Independence

Independence and Exclusive

Difference between Mutually exclusive and independent events.

Mutually exclusive events Independent events

1 When the occurrence is not When the occurrence of one
simultaneous for two events event does not control the
then they are termed as occurrence of the other event,
Mutually exclusive events. they are independent.
2 The non-occurrence of an There is no influence of an
event will end up in the occurrence with another and .
occurrence of an event they are independent of each
3 The mathematical formula for other. The mathematical formula for
mutually exclusive events can be independent events can be
represented as P(A ∩ B) = 0 defined as P(A ∩ B) = P(A).P(B)
4 The sets will not overlap in the The sets may overlap in the case
case of mutually exclusive events. of independent events.

Amit (ISI, Chennai) ProbDist July 27, 2024 35 / 47

 Conditional Probability Bayes’ Law

Contents

1 Introduction
History
Random Events
Variables & Outcome
Measuring Probability

2 Axiom
Probability Types
Probability Rules

3 Conditional Probability
Independence
Bayes’ Law
Conditional Problems

Amit (ISI, Chennai) ProbDist July 27, 2024 36 / 47

 Conditional Probability Bayes’ Law

Bayes’ Theorem
Bayes’ rule provides a method to find out conditional probability, with
condition and event reversed.

Amit (ISI, Chennai) ProbDist July 27, 2024 37 / 47

 Conditional Probability Bayes’ Law

Bayes’ Theorem
Bayes’ rule provides a method to find out conditional probability, with
condition and event reversed.

It’s deceptively simple formula. The formal definition of the rule is :

P(B|A)P(A)
P(A|B) =
P(B)
We are interested in finding out a patient’s probability of having COVID, if
they are a COPD patient already. ‘Being COPD’ may complicate COVID.

Amit (ISI, Chennai) ProbDist July 27, 2024 37 / 47

 Conditional Probability Bayes’ Law

Bayes’ Theorem
Bayes’ rule provides a method to find out conditional probability, with
condition and event reversed.

It’s deceptively simple formula. The formal definition of the rule is :

P(B|A)P(A)
P(A|B) =
P(B)
We are interested in finding out a patient’s probability of having COVID, if
they are a COPD patient already. ‘Being COPD’ may complicate COVID.

A could mean the event ‘Patient is COVID +ve’. Past data tells you that 10%
of patients entering your clinic have COVID. P(A) = 0.10.

Amit (ISI, Chennai) ProbDist July 27, 2024 37 / 47

 Conditional Probability Bayes’ Law

Bayes’ Theorem
Bayes’ rule provides a method to find out conditional probability, with
condition and event reversed.

It’s deceptively simple formula. The formal definition of the rule is :

P(B|A)P(A)
P(A|B) =
P(B)
We are interested in finding out a patient’s probability of having COVID, if
they are a COPD patient already. ‘Being COPD’ may complicate COVID.

A could mean the event ‘Patient is COVID +ve’. Past data tells you that 10%
of patients entering your clinic have COVID. P(A) = 0.10.

B could mean that ‘Patient is COPD. Five percent of the clinic’s patients are
COPD. P(B) = 0.05.

Amit (ISI, Chennai) ProbDist July 27, 2024 37 / 47

 Conditional Probability Bayes’ Law

Bayes’ Theorem
Bayes’ rule provides a method to find out conditional probability, with
condition and event reversed.

It’s deceptively simple formula. The formal definition of the rule is :

P(B|A)P(A)
P(A|B) =
P(B)
We are interested in finding out a patient’s probability of having COVID, if
they are a COPD patient already. ‘Being COPD’ may complicate COVID.

A could mean the event ‘Patient is COVID +ve’. Past data tells you that 10%
of patients entering your clinic have COVID. P(A) = 0.10.

B could mean that ‘Patient is COPD. Five percent of the clinic’s patients are
COPD. P(B) = 0.05.

You might also know that among those patients diagnosed COVID, 7% are
COPD. This is your B|A : the probability that a patient is COPD, given that
they have COVID, is 0.07.
Amit (ISI, Chennai) ProbDist July 27, 2024 37 / 47
 Conditional Probability Bayes’ Law

Bayes’ Theorem
Bayes’ rule provides a method to find out conditional probability, with
condition and event reversed.

It’s deceptively simple formula. The formal definition of the rule is :

P(B|A)P(A)
P(A|B) =
P(B)
We are interested in finding out a patient’s probability of having COVID, if
they are a COPD patient already. ‘Being COPD’ may complicate COVID.

A could mean the event ‘Patient is COVID +ve’. Past data tells you that 10%
of patients entering your clinic have COVID. P(A) = 0.10.

B could mean that ‘Patient is COPD. Five percent of the clinic’s patients are
COPD. P(B) = 0.05.

You might also know that among those patients diagnosed COVID, 7% are
COPD. This is your B|A : the probability that a patient is COPD, given that
they have COVID, is 0.07.
Amit (ISI, Chennai) ProbDist July 27, 2024 37 / 47
 Conditional Probability Bayes’ Law

Bayes’ Theorem

Bayes’ theorem tells you: P(A|B) = (0.07 ∗ 0.1)/0.05 = 0.14.

If one is a COPD patient then chances that s/he maybe COVID +ve is 14%.

Amit (ISI, Chennai) ProbDist July 27, 2024 38 / 47

 Conditional Probability Bayes’ Law

Bayes’ Theorem

Bayes’ theorem tells you: P(A|B) = (0.07 ∗ 0.1)/0.05 = 0.14.

If one is a COPD patient then chances that s/he maybe COVID +ve is 14%.
P(B|A)P(A) P(A)
Let’s look at the simple formula again. . . P(A|B) = P(B) = P(B|A) P(B)

Amit (ISI, Chennai) ProbDist July 27, 2024 38 / 47

 Conditional Probability Bayes’ Law

Bayes’ Theorem

Bayes’ theorem tells you: P(A|B) = (0.07 ∗ 0.1)/0.05 = 0.14.

If one is a COPD patient then chances that s/he maybe COVID +ve is 14%.
P(B|A)P(A) P(A)
Let’s look at the simple formula again. . . P(A|B) = P(B) = P(B|A) P(B)
P(B|A)P(A) P(B|A)P(A)
Rewrite P(B) = P(A∩B)+P(Ac ∩B)

Amit (ISI, Chennai) ProbDist July 27, 2024 38 / 47

 Conditional Probability Bayes’ Law

Bayes’ Theorem

Bayes’ theorem tells you: P(A|B) = (0.07 ∗ 0.1)/0.05 = 0.14.

If one is a COPD patient then chances that s/he maybe COVID +ve is 14%.
P(B|A)P(A) P(A)
Let’s look at the simple formula again. . . P(A|B) = P(B) = P(B|A) P(B)
P(B|A)P(A) P(B|A)P(A)
Rewrite P(B) = P(A∩B)+P(Ac ∩B)

P(B|A)P(A)
P(B|A)P(A)+P(B|Ac )P(Ac )

Amit (ISI, Chennai) ProbDist July 27, 2024 38 / 47

 Conditional Probability Bayes’ Law

Bayes’ Theorem

Bayes’ theorem tells you: P(A|B) = (0.07 ∗ 0.1)/0.05 = 0.14.

If one is a COPD patient then chances that s/he maybe COVID +ve is 14%.
P(B|A)P(A) P(A)
Let’s look at the simple formula again. . . P(A|B) = P(B) = P(B|A) P(B)
P(B|A)P(A) P(B|A)P(A)
Rewrite P(B) = P(A∩B)+P(Ac ∩B)

P(B|A)P(A)
P(B|A)P(A)+P(B|Ac )P(Ac )

P(B|A) is what we have, P(A|B) is what we desire to obtain.

Amit (ISI, Chennai) ProbDist July 27, 2024 38 / 47

 Conditional Probability Bayes’ Law

Bayes’ Theorem

1 A couple has 2 children and the older child is a boy. If the probabilities of
having a boy or a girl are both 50%, what’s the probability that the
couple has two boys? We already know that the older child is a boy. The
probability of two boys is equivalent to the probability that the younger
child is a boy, which is 50%.
2 A couple has two children, of which at least one is a boy. If the
probabilities of having a boy or a girl are both 50%, what is the
probability that the couple has two boys?

Amit (ISI, Chennai) ProbDist July 27, 2024 39 / 47

 Conditional Probability Bayes’ Law

Bayes’ Theorem

1 A couple has 2 children and the older child is a boy. If the probabilities of
having a boy or a girl are both 50%, what’s the probability that the
couple has two boys? We already know that the older child is a boy. The
probability of two boys is equivalent to the probability that the younger
child is a boy, which is 50%.
2 A couple has two children, of which at least one is a boy. If the
probabilities of having a boy or a girl are both 50%, what is the
probability that the couple has two boys?

Same question?

Amit (ISI, Chennai) ProbDist July 27, 2024 39 / 47

 Conditional Probability Bayes’ Law

Bayes’ Theorem

1 A couple has 2 children and the older child is a boy. If the probabilities of
having a boy or a girl are both 50%, what’s the probability that the
couple has two boys? We already know that the older child is a boy. The
probability of two boys is equivalent to the probability that the younger
child is a boy, which is 50%.
2 A couple has two children, of which at least one is a boy. If the
probabilities of having a boy or a girl are both 50%, what is the
probability that the couple has two boys?

Same question? Possibly

Amit (ISI, Chennai) ProbDist July 27, 2024 39 / 47

 Conditional Probability Bayes’ Law

Bayes’ Theorem

1 A couple has 2 children and the older child is a boy. If the probabilities of
having a boy or a girl are both 50%, what’s the probability that the
couple has two boys? We already know that the older child is a boy. The
probability of two boys is equivalent to the probability that the younger
child is a boy, which is 50%.
2 A couple has two children, of which at least one is a boy. If the
probabilities of having a boy or a girl are both 50%, what is the
probability that the couple has two boys?

Same question? Possibly not!

We know that one is a boy, so the only question is whether the other one is a
boy, and the chances of that being the case are 50%. So again, the answer is
50%

Amit (ISI, Chennai) ProbDist July 27, 2024 39 / 47

 Conditional Probability Bayes’ Law

Bayes’ Theorem

1 A couple has 2 children and the older child is a boy. If the probabilities of
having a boy or a girl are both 50%, what’s the probability that the
couple has two boys? We already know that the older child is a boy. The
probability of two boys is equivalent to the probability that the younger
child is a boy, which is 50%.
2 A couple has two children, of which at least one is a boy. If the
probabilities of having a boy or a girl are both 50%, what is the
probability that the couple has two boys?

Same question? Possibly not!

We know that one is a boy, so the only question is whether the other one is a
boy, and the chances of that being the case are 50%. So again, the answer is
50%

This makes perfect sense.

Amit (ISI, Chennai) ProbDist July 27, 2024 39 / 47

 Conditional Probability Bayes’ Law

Bayes’ Theorem

1 A couple has 2 children and the older child is a boy. If the probabilities of
having a boy or a girl are both 50%, what’s the probability that the
couple has two boys? We already know that the older child is a boy. The
probability of two boys is equivalent to the probability that the younger
child is a boy, which is 50%.
2 A couple has two children, of which at least one is a boy. If the
probabilities of having a boy or a girl are both 50%, what is the
probability that the couple has two boys?

Same question?

We know that one is a boy, so the only question is whether the other one is a
boy, and the chances of that being the case are 50%. So again, the answer is
50%

This makes perfect sense. It also happens to be incorrect.

Amit (ISI, Chennai) ProbDist July 27, 2024 39 / 47

 Conditional Probability Bayes’ Law

Bayes’ Theorem

1 A couple has 2 children and the older child is a boy. If the probabilities of
having a boy or a girl are both 50%, what’s the probability that the
couple has two boys? We already know that the older child is a boy. The
probability of two boys is equivalent to the probability that the younger
child is a boy, which is 50%.
2 A couple has two children, of which at least one is a boy. If the
probabilities of having a boy or a girl are both 50%, what is the
probability that the couple has two boys?

Same question?

We know that one is a boy, so the only question is whether the other one is a
boy, and the chances of that being the case are 50%. So again, the answer is
50%

This makes perfect sense. It also happens to be incorrect.

Amit (ISI, Chennai) ProbDist July 27, 2024 39 / 47

 Conditional Probability Conditional Problems

Contents

1 Introduction
History
Random Events
Variables & Outcome
Measuring Probability

2 Axiom
Probability Types
Probability Rules

3 Conditional Probability
Independence
Bayes’ Law
Conditional Problems

Amit (ISI, Chennai) ProbDist July 27, 2024 40 / 47

 Conditional Probability Conditional Problems

Two-children family problem

1 A couple has two children, the older of which is a boy. What is the
probability that they have two boys?
2 A couple has two children, one of which is a boy. What is the probability
that they have two boys?

Amit (ISI, Chennai) ProbDist July 27, 2024 41 / 47

 Conditional Probability Conditional Problems

Two-children family problem

1 A couple has two children, the older of which is a boy. What is the
probability that they have two boys?
2 A couple has two children, one of which is a boy. What is the probability
that they have two boys?

Define three events A, B and C as follows:

Amit (ISI, Chennai) ProbDist July 27, 2024 41 / 47

 Conditional Probability Conditional Problems

Two-children family problem

1 A couple has two children, the older of which is a boy. What is the
probability that they have two boys?
2 A couple has two children, one of which is a boy. What is the probability
that they have two boys?

Define three events A, B and C as follows:

A = both children are boys
B = the older child is a boy
C = one of their children is a boy.

Amit (ISI, Chennai) ProbDist July 27, 2024 41 / 47

 Conditional Probability Conditional Problems

Two-children family problem

1 A couple has two children, the older of which is a boy. What is the
probability that they have two boys?
2 A couple has two children, one of which is a boy. What is the probability
that they have two boys?

Define three events A, B and C as follows:

A = both children are boys
B = the older child is a boy
C = one of their children is a boy.

Question 1 is asking for P(A|B) and Question 2 is asking for P(A|C ). The first
use simple version of Bayes’.

P(A)P(B|A) 0.25.1 1
P(A|B) = = =
P(B) 0.5 2
To find P(A|C ), we need to find P(C ), the prior probability that the couple
has at least one boy is 1 − P(both children are girls) = 1 − 14 = 34
Amit (ISI, Chennai) ProbDist July 27, 2024 41 / 47
 Conditional Probability Conditional Problems

Conditional Probability Problem—2

For three events A, B, and C , we know that
A and C are independent,
B and C are independent,
A and B are disjoint

Amit (ISI, Chennai) ProbDist July 27, 2024 42 / 47

 Conditional Probability Conditional Problems

Conditional Probability Problem—2

For three events A, B, and C , we know that
A and C are independent,
B and C are independent,
A and B are disjoint and that
P(A ∪ C ) = 23 , P(B ∪ C ) = 34 , P(A ∪ B ∪ C ) = 11
12 .

Find P(A), P(B) and P(C ).

Amit (ISI, Chennai) ProbDist July 27, 2024 42 / 47

 Conditional Probability Conditional Problems

Conditional Probability Problem—2

For three events A, B, and C , we know that
A and C are independent,
B and C are independent,
A and B are disjoint and that
P(A ∪ C ) = 23 , P(B ∪ C ) = 34 , P(A ∪ B ∪ C ) = 11
12 .

Find P(A), P(B) and P(C ).

Solution : Draw venn diagram, write down the probability equations, solve
equations!

Amit (ISI, Chennai) ProbDist July 27, 2024 42 / 47

 Conditional Probability Conditional Problems

Conditional Probability Problem—2

For three events A, B, and C , we know that
A and C are independent,
B and C are independent,
A and B are disjoint and that
P(A ∪ C ) = 23 , P(B ∪ C ) = 34 , P(A ∪ B ∪ C ) = 11
12 .

Find P(A), P(B) and P(C ).

Solution : Draw venn diagram, write down the probability equations, solve
equations!

Amit (ISI, Chennai) ProbDist July 27, 2024 42 / 47

 Conditional Probability Conditional Problems

Conditional Probability Problem—3

In certain town, it’s rainy one third of the days. Given that it is rainy, there
will be heavy traffic with probability frac12, and given that it is not rainy,
there will be heavy traffic with probability 14 . If it’s rainy and there is heavy
traffic, I arrive late for work with probability 12 . On the other hand, the
probability of being late is reduced to 81 if it is not rainy and there is no heavy
traffic. In other situations (rainy and no traffic, not rainy and traffic) the
probability of being late is 0.25. You pick a random day.
What is the probability that it’s not raining and there is heavy traffic and
I am not late?
What is the probability that I am late?
Given that I arrived late at work, what is the probability that it rained
that day?

Amit (ISI, Chennai) ProbDist July 27, 2024 43 / 47

 Conditional Probability Conditional Problems

Conditional Probability Problem—4

A box contains three coins: two regular coins and one fake two-headed coin
(P(H) = 1),
You pick a coin at random and toss it. What is the probability that it
lands heads up?
You pick a coin at random and toss it, and get heads. What is the
probability that it is the two-headed coin?

Amit (ISI, Chennai) ProbDist July 27, 2024 44 / 47

 Conditional Probability Conditional Problems

Conditional Probability Problem—4

This is another typical problem for which the law of total probability is useful.
Let C1 be the event that you choose a regular coin, and let C2 be the event
that you choose the two-headed coin. Note that C1 and C2 form a partition of
the sample space. We already know that
P(H|C1 ) = 0.5, P(H|C2 ) = 1.
Thus, we can use the law of total probability to write
= P(H|C1 )P(C1 ) + P(H|C2 )P(C2 ) (1)
1 2 1
= . + 1. (2)
2 3 3
(3)
Now, for the second part of the problem, we are interested in P(C2 |H). We use
Bayes’ rule
P(H|C2 )P(C2 )
P(C2 |H) = (4)
P(H)
1. 13
= (5)
23
Amit (ISI, Chennai) ProbDist July 27, 2024 45 / 47
 Conditional Probability Conditional Problems

Tailpiece : A trading problem

A company purchases copper concentrate from international trading

companies, the payment to the supplier is based on chemical analysis by
both the trader and the recipient. The chemical analysis evaluates the
percentage of copper, silver and gold, which forms the basis of payment as
long as they don’t differ by a specified margin called the split limit.

Amit (ISI, Chennai) ProbDist July 27, 2024 46 / 47

 Conditional Probability Conditional Problems

Tailpiece : A trading problem

A company purchases copper concentrate from international trading

companies, the payment to the supplier is based on chemical analysis by
both the trader and the recipient. The chemical analysis evaluates the
percentage of copper, silver and gold, which forms the basis of payment as
long as they don’t differ by a specified margin called the split limit.

In which case the samples are sent to ‘independent’ umpires and payment
is based on umpires’ analysis.

Amit (ISI, Chennai) ProbDist July 27, 2024 46 / 47

 Conditional Probability Conditional Problems

Tailpiece : A trading problem

A company purchases copper concentrate from international trading

companies, the payment to the supplier is based on chemical analysis by
both the trader and the recipient. The chemical analysis evaluates the
percentage of copper, silver and gold, which forms the basis of payment as
long as they don’t differ by a specified margin called the split limit.

In which case the samples are sent to ‘independent’ umpires and payment
is based on umpires’ analysis.

The company often finds the trader’s analysis is higher than their lab
analysis and occasionally exceeding the limit as well, forcing them to send
samples to the international umpires at a staggering cost. Apart from
paying a perceived higher amount.

Amit (ISI, Chennai) ProbDist July 27, 2024 46 / 47

 Conditional Probability Conditional Problems

Tailpiece : A trading problem

A company purchases copper concentrate from international trading

companies, the payment to the supplier is based on chemical analysis by
both the trader and the recipient. The chemical analysis evaluates the
percentage of copper, silver and gold, which forms the basis of payment as
long as they don’t differ by a specified margin called the split limit.

In which case the samples are sent to ‘independent’ umpires and payment
is based on umpires’ analysis.

The company often finds the trader’s analysis is higher than their lab
analysis and occasionally exceeding the limit as well, forcing them to send
samples to the international umpires at a staggering cost. Apart from
paying a perceived higher amount.

Amit (ISI, Chennai) ProbDist July 27, 2024 46 / 47

 Conditional Probability Conditional Problems

A trading problem

So a loss function model was fitted to the data and after taking into
account the variability of lab and trader and the actual lab analysis of the
particular lot, what should be declared for payment was decided.

Amit (ISI, Chennai) ProbDist July 27, 2024 47 / 47

 Conditional Probability Conditional Problems

A trading problem

So a loss function model was fitted to the data and after taking into
account the variability of lab and trader and the actual lab analysis of the
particular lot, what should be declared for payment was decided.

To facilitate the implementation a spread-sheet was prepared to take the

inputs and suggest the declared value to the company.

Amit (ISI, Chennai) ProbDist July 27, 2024 47 / 47