Probability

What is probability?
Possibly,it would rain tonight
There is a high chance of my getting the job next
month
Here the words like possibly, high chance indicates a
degree of uncertainty about the happening of the event.
All these events are random in nature.
Prob. Theory includes studying such events to make
the best possible decision in the face of uncertainties.
Probability is the science of decision – making with
calculated risks in the face of uncertainty.
Basic terminology
Random experiment : it is a procedure that can result
in many outcomes such that although all outcomes
may be known it is not possible to predict the outcome
associated with a single experiment. E.g.
Tossing of a coin whose all possibilities are known,
head and tail, but at each particular toss, we don’t
know if it will face head or tail.
Sample space : of a random experiment is the set S
that includes all possible outcomes of an experiment.
In tossing of a fair coin the set of all the possible
outcomes is {H,T}
Here, {H,T} is the sample space for this experiment.
Questions
Suppose a committee of 2 is to selected from a group
consisting of 5 people say R,S,V,A,N. find the sample
space.
A manufacturer purchases equipments for three vendor
A,B,C. Let (1,2) denote the event that on two
successive days, the first day order goes to vendor A
and on the second day vendor B gets the order. Write
the sample space.
Trail : each experiment is known as a trial. E.g.
tossing a coin once, is a single trial.
Event : outcomes of the trials are known as events.
E.g. in tossing of a coin, outcomes are either H or T.
so, H and T are the events.
Rules of event operations
Intersection of events
Union of events
Complement of events
Types of events
Exhaustive events
Mutually exclusive events
Independent and dependent events
Equally likely events
Complementary event
Probability assigning techniques
Classical technique
Relative frequency technique
Subjective approach
Questions
A box contains 2 red, 5 white and 6 blue balls. What is
the prob. That out of two balls drawn, one is white and
one is blue?
What is the prob. That a leap year selected at random
will contain 53 sundays?
A box contains 8 good pens, 4 pens with minor defects
and 3 pens with major defects. Four pens are picked at
random without replacement. Find the prob. That
All pens are defective
Atleast one pen is good
Exactly 2 pens are good
One pen is good, one has a minor defect and 2 have a
major defect
One pen is good and 3 have major defects
Types of probability
Marginal probability
Union probability
Joint probability
Conditional probability
Laws and theorems of probability
Additive law
Multiplication law
Theory of independence
Theory of independence

Independence : A and B are independent if and only if

P(A&B)=P(A)*P(B)
Laws and theorems of probability
Additive law
Conditional probability
Multiplication law
Theory of independence
Baye’s theorem
Theory of independence

Independence : A and B are independent if and only if

P(A&B)=P(A)*P(B)
EXAMPLE
 ABRC, a leading marketing research firm in India, wants to
collect information about households with computers and
Internet access in urban Mumbai. After conducting an
intensive survey, it was revealed that 60% of the
households have computers with Internet access; 70% of
the household have two or more computer sets. Suppose
50% of the households have computers with Internet
connection and two or more computers. A household with
computer is randomly selected.
What is the probability that the household has computers with
Internet access or two or more computers?
What is the probability that the household has computers with
Internet access or two or more computers, but not both?
What is the probability that the household has neither
computers with Internet access nor two or more computers?
Example
 A company is interested in understanding the consumer behaviour of the
capital of the newly formed state Chhattisgarh, that is, Raipur. For this
purpose, the company has selected a sample of 300 consumers and asked a
simple question, ”Do you enjoy shopping?” out of 300 respondents, 200
were males and 100 were females. Out of 200 males, 120 responded Yes
and out of 100 females, 70 responded yes. A respondent is selected
randomly. Construct a probability matrix and ascertain the probability that:
 The respondent is a male
 Enjoys shopping
 Is a female and enjoys shopping
 Is a male and does not enjoy shopping
 Is a female or enjoys shopping
 Is a male or does not enjoy shopping
 Is a male or female
EXAMPLE
A market research firm is interested in surveying certain
attitudes in a small community. There are 1250
households broken down acc to income, ownership of a
telephone and ownership of a T.V.

Household with annual Household with annual

income <=30000 income >30000

Telephone No telephon Telephone No telephon

subscr. subscr.
Own TV set 270 200 180 100

No TV set 180 100 120 100

What is the prob of obtaining a TV owner in drawing
of households at random?
If a household has annual income over Rs. 30,000 and
is a telephone subscriber, what is the prob that he has a
TV?
What is the conditional prob of drawing a household
that owns a TV, given that the household is a
telephonic subscriber?
Practice problem

If HIV has a prevalence of 3% in San Francisco, and a

particular HIV test has a false positive rate of .001 and
a false negative rate of .01, what is the probability that
a random person selected off the street will test
positive?
 Answer
Conditional probability: the
probability of testing + given that Joint probability of being + and
a person is + testing +
Marginal probability of carrying
the virus.
P(test +)=.99 P (+, test +)=.0297

P(+)=.03 P(test - )= .01

P(+, test -)=.003
P(test +) = .001 P(-, test +)=.00097
P(-)=.97

P(test -) = .999 P(-, test -) = .96903

______________
1.0
Marginal probability of testing
positive

P(test +)=.0297+.00097=.03067
P(+&test+)P(+)*P(test+)
.0297 .03*.03067 (=.00092)

 Dependent!
 Law of total probability

P(test  )  P(test  /HIV)P(HIV )  P(test  /HIV)P(HIV-)

One of these has to be true (mutually exclusive,

collectively exhaustive). They sum to 1.0.

P(test  )  .99(.03)  .001(.97)

 Law of total probability

 Formal Rule: Marginal probability for event A=

P(A)  P(A | B1 ) P(B1 )  P(A | B2 ) P(B2 )    P(A | Bk ) P(Bk )

k
 Where: Bi  1.0 and P(Bi &B j )  0 (mutually exclusive)
i 1
Bayes’ rule
Bayes’ Rule: derivation
Definition:
Let A and B be two events with P(B)  0. The
conditional probability of A given B is:

P( A & B)
P( A / B) 
P( B)
Bayes’ Rule: derivation
P( A & B)
P( A / B) 
P( B)

can be re-arranged to:

P( A & B)  P( A / B) P( B)
and, since also:
P( A & B)
P( B / A)   P ( A & B)  P ( B / A) P( A)
P( A)

P ( A / B ) P( B )  P( A & B)  P( B / A) P( A)
P ( A / B ) P( B )  P( B / A) P( A)
P( B / A) P ( A)
 P( A / B) 
P( B)
 Bayes’ Rule:

P( B / A) P( A)
P( A / B) 
P( B)
OR

P( B / A) P( A)
P( A / B)  From the
P( B / A) P( A)  P( B / ~ A) P(~ A)
“Law of Total
Probability”
Bayes’ Rule:
Why do we care??
Why is Bayes’ Rule useful??
It turns out that sometimes it is very useful to be able
to “flip” conditional probabilities. That is, we may
know the probability of A given B, but the probability
of B given A may not be obvious. An example will
help…
In-Class Exercise
If HIV has a prevalence of 3% in San Francisco, and a
particular HIV test has a false positive rate of .001 and
a false negative rate of .01, what is the probability that
a random person who tests positive is actually infected
(also known as “positive predictive value”)?
 Answer: using probability treeP(test +)=.99 P (+, test +)=.0297

 
P(+)=.03
P(test - = .01)

  P(+, test -)=.003

P(test +) = .001
P(-, test +)=.00097
P(-)=.97

  P(-, test -) = .96903

 
P(test -) = .999
  ______________
1.0

A positive test places one on either of the two “test +” branches.

But only the top branch also fulfills the event “true infection.”
Therefore, the probability of being infected is the probability of being on the top
branch given that you are on one of the two circled branches above.
P (test  &true  ) .0297
P (  / test )    96.8%
P (test  ) .0297  .00097
 Answer: using Bayes’ rule
 

 
P(test  / true ) P (true )
P(true  / test  )  
P(test  / true ) P (true )  P (test  / true) P(true)
.99(.03)
 96.8%
.99(.03)  .001(.97)
 
 
 
Practice problem
An insurance company believes that drivers can be
divided into two classes—those that are of high risk and
those that are of low risk. Their statistics show that a
high-risk driver will have an accident at some time
within a year with probability .4, but this probability is
only .1 for low risk drivers.
a) Assuming that 20% of the drivers are high-risk, what
is the probability that a new policy holder will have
an accident within a year of purchasing a policy?
b) If a new policy holder has an accident within a year
of purchasing a policy, what is the probability that he
is a high-risk type driver?
EXAMPLE
In a basin area where oil is likely to be found
underneath the surface, there are three locations with
three different types of earth composition, say C1, C2,
C3. the probability for these three compositions are
0.5, 0.3 and 0.2, respectively. Further, it has been
found from the past experience that the probabilities of
finding oil due to drilling of well at these locations is
0.2, 0.4 and 0.3, respectively. Suppose a well is drilled
at a location, and it yields oil, what is the probability
the earth composition was C2?
EXAMPLE
 The foreman in an iron foundry knows that the sand used for
moulding iron castings is too dry 5% of the time and too wet
2% of the time. He also knows that the defective castings
occur 0.5% of the time when the sand has the correct amount
of moisture; 5% of the time when the sand is too dry; and
25% of the time when the sand is too wet. Suppose a casting
is selected at random from the batch just produced and is
found to be defective. What is the probability that each of the
following is true?
The sand was too wet
The sand was too dry
The sand had the correct amount of moisture.
EXAMPLE
The compressors used in refrigerators are
manufactured by XYZ Co. at three factories at Pune,
Nasik and Nagpur. It is known that the Pune factory
produces twice as many compressors as the Nasik one,
which produces the same number as the Nagpur one.
Experience also shows that 0.2% of the compressors
produced at Pune and Nasik are defective and so are
0.4% of those produced at Nagpur.
A quality control engineer while maintaining a
refrigerator finds a defective compressor. What is the
probability that the Nasik factory is not be blamed.