UNIT 2 THEOREMS PROBABILITY

Structure
2.1 Introduction
Objectives
2.2 Some Elementary Theorems
2.3 General Addition Rule
2.4 Conditional Probability and Independence
2.4.1 Conditional Probability
2.4.2 Independent Events and Multiplication Rule
2.4.3 Theorem of Total Probability and Bayes Theorem
2.5 Summary
2.1 INTRODUCTION
You have already learnt about probability axioms and ways to evaluate probability
of events in some simple cases.
In this unit, we discuss ways to evaluate the probability of combination of events.
For this, we derive the addition rule which deals with the probability of union of
two events and the multiplication rule which deals with thc probability of
intersection of two events. Two important concepts namely : Conditional
Probability and independence of events, are introduced and Bayes theorem, which
deals with conditional probability is presented.
Objectives
After reading this unit, you should be able to
* evaluate the probability of certain combination of events involving union,
intersection and complementation,
* evaluate conditional probability,
* check independence of two or more events, and
* apply Bayes theorem to find the probability that the "effectY'A was
"caused" by the event B.
2.2 SOME ELEMENTARY THEOREMS
Recall the axiomatic definition of probability which you have read in Section 1.4.
Using these axioms of probability, it is possible to derive many results which are
very useful in applications. We present some of these results in this section.
Theorem 1 :
If 4 is the empty set then
P(4) b= 0
Proof :
For any eventA, A a A U 9. Also A and 9 are mutually exclusive as
A f l 4 = 4. Hence by Axiom 3 of probability axioms,
Probability Concepts
which implies that P ( $ ) - 0.
Theorem 2 :
<
I ~ A is the complementary event of A, then
P ( A ) - I - P ( A )
Proof:
Note that S - A U A where S is the sample space. FurtherA a n d 2 are
mutually exclusive asA n - $. Therefore by Axiom 3 of probability axioms
P ( S ) - P( AUA) - P ( A ) + P ( A )
(2.1)
But by Axiom 2, P ( S ) - 1 and hence equation ( 2.1 ) gives
P ( A ) + P ( A ) - 1
i.e. P ( A ) = I - P ( A )
I .L-rem 3 :
If A and B are any two events then
P ( A U B ) - P ( A ) + P ( B ) - p ( A n B f
Proof :
The below given Venn diagram (Figure 2.1) suggests that eventsA U B and B
can be expressed as
A n B ~ n x
B - ( A ~ B ) u ( B ~ A )
A UB - A U ( B ~ A )
But ( A n B ) and ( B n A ) are mutually exclusive and hence by Axiom
equation (2.2) gives
P ( B ) - P ( ( A ~ B ) u ( B ~ ~ ) ) - P ( A ~ B ) + P ( B ~ ~ )
Similarly equation (2.3) gives
P ( A U B ) - P ( A U ( B ~ ~ ) ) - P ( A ) + P ( B ~ A )
From equation (2.4) and equation (2.5),
P ( A U B ) - P ( A ) + P ( B ) - P ( A n B )
Remark 1 :
Theformul aP(AUB) - P( A) +P( B) - P( AnB) i s c al l e dt he addi t wn
rule and is used to evaluate the probability of occurrence of at least one of the
two given events. You will learn the general addition rule in Section 2.3.
Theorem 4 :
IfA andBaretwoeventssuchthatACB, then P ( A ) s P ( B ) ;
Proof:
As A C B, we can express B as the union of two mutually exclusive events
A and ( B fi A ) ( Ref. Figure 2.2) i.e.
Figme 2 2
B = A U ( B ~ ~ )
Axiom 3, together with equation ( 2.6 ) gives
P ( B ) - P ( ' A ) + P ( B ~ ~ ) (2.7)
But P ( B n A ) r 0 (by Axioml) and hence equation (2.7) implies
P ( B ) r P ( A ) .
~ o t e t h a t e ~ u a t i o n ( 2 . 7 ) a l s o i m~ l i e s ~ ( ~ ~ ~ ) - P ( B ) - P ( A ) , if A CB.
Sometimes we denote the event B fl A by the set theoretic difference B - A and
I
i
so we can write equation (2.7) as
P ( B - A ) = P ( B ) - P ( A )
,
i
provided A C B.
1 Example 1 :
i Suppose that A and B are events for which
1 P ( A ) = x , P( B) = y and P ( A n B ) = z. Express P ( ~ u B ) and
I
From De-Morgan's laws it is known that ( A U B ) = ( A n B ) and therefore
Theorems of Probability
~ l s o , from figue 2.1, ( An B ) = B- ( A n B ).
I
= P ( B ) - P ( A n B ) (using equation (2.8) as
( A n B ) C B )
, El
d
For the data of Example 1, find
p ( ; i u ~ ) , ~ ( ; i n i i ) and ~ [ ( ~ n i i ) u ( ~ n ; i ) ]
The probability that a person stopping at a gas station will ask to have his tyres
checked is 0.12, the probability that he will ask to have his oil checked is 0.29,
and the probability that he will ask them both checked is 0.07. Find the
probability that a person stopping at the gas station will have (i) his tyres or his
oil checked and (ii) will have neither his tyres nor his oil checked.
2.3 GENERAL ADDITION RULE
You have already learnt the addition rule ( Theorem 3 ) for finding the probability
of occurrence of at least one of the two given events. In this section, you will learn
the general addition rule which will help you to evaluate the probability of
occurrence of at least one of the given n events. If the a events are denoted by
A1 , A2 ,..An then our aim is to find a formula for evaluating
P (A1 U A2 U ... U An ) . For this, let us first take 3 eventsA , B and C and use
Theorem 3 to find the expression for P (A U B U C ) .
Theorem 4 :
If A , B and C are any three events, then
P ( A U B U C ) = P ( A ) + P ( B ) + P ( C ) - P ( A n B )
i
Proof :
L
1 Let us denote B U C by D and therefore
where equation (2.9) follows because of Theorem 3. Also using Theorem 3
t again,
P ( D ) = P ( B U C ) - P ( B ) + P ( C ) - P ( B n C )
and
Substituting for P ( D ) and P ( A n D ) in equation (2.9) we obtain
which proves Theorem 4.
Theorem 4 above, suggests the form of the expression for
P ( A1 U A2 U ... U An ) and this we present in Theorem 5, called the general
addition rule.
Theorem 5 :
LetAl,A2, ... An be any n events. Then
( -1
P ( AI f l A2 f l ... n An )
(2.12)
The proof of Theorem 5 can be given by using the principle of mathematical
induction.
Example 2 :
Suppose that in the maintenance of a large medical records file for insurance
purposes the probability of an error in processing is 0.0010, the probability of
'
an error in filling is 0.0009, the probability of an error in retrieving is 0.0012,
the probability of an error in processing as well as filling is 0.0002, the
probability of an error in processing as well as retrieving is 0.0003, the
probability of an error in filling as well as retrieving is 0.0003, and probability
of an error in processing, filling as well as retrieving is 0.0001. What is the
probability of making at least one of these errors ?
Tbcorem~ of Probability
Solution :
LetA , B and C denote respectively the events, error in processing, error in
filling and error in retrieving. From the given data,
P ( A ) = 0.0010, P ( B ) = 0.0009, P ( C ) = 0.0012
P ( A n B ) = 0. 0002, P( AnC) = 0. 0003, P( BnC) = 0.0003
P ( A n B n c ) = o.oooi.
The required probability is:
P ( A U B U C ) = P ( A ) + P ( B ) + P ( C ) - P ( A n B ) - P ( A n c )
- P ( B n C ) + P ( A n B n C ) .
= 0.0010 + 0.0009 + 0.0012 - 0.0002 - 0.0003 - 0.0003 + O.O(
= 0.0024.
E3
Suppose thatA , B and C are events such that
1
P ( A ) = P ( B ) = P ( C ) = -,
4
P ( A n B ) = P ( C n B ) = 0, and P ( A n C ) Evaluate the
= 3
probability that at least one of the eventsA , B or C occur.
E4
Show that for any 3 eventsA , B and C
P ( A U B U C ) 4 P ( A ) + P ( B ) + P ( C ) .
2.4 CONDITIONAL PROBABILITY AND
INDEPENDENCE
Recall the definition of probability and note that it is meaningful to ask for the
probability of an event only if we refer to a given sample space S. Thus the symbol
P ( A ) for the probability of A really means the probability ofA given some sample
space S. Since there are situations where we are interested in probabilities of A with
respect to more than one sample spaces, the notation P ( A/ S ) is introduced and it
i
is read as "The conditional probability ofA relative to the sample space S" or "the
conditional Probability of A given the sample space 5"'. We continue using
P (A ) for P ( A/S ) , whenever the choice of S is clearly understood. Note that the
sample space S itself is an event and therefore you may be interested in generalizing
the notion of P (A/S ) to P ( A/B ) or P ( B/A ) whenA , Bare events of S. The
symbol P ( B/A ) is read as "the conditional probability of the event B relative to
the event A " or "the conditional probability of the event B given that eventA has
occurred". This notion and other details we discuss in the subsequent sections.
24.1 Conditional probability
When we evaluate P ( B ) we are asking ourselves how probable it is that we shall
be in B, knowing that we must lie in S. This follows because P ( B ) really refers to
I
P ( B/S ). But when we compute P ( B/A ) , we are asking ourselves h w probable
it is that we shall be in B , knowing that we must be in A. Thus when we
t
computing P ( B/A ) we are asssentially computing P ( B ) with respect to the
reduced sample spaceA , rather than with respect to the original sample space S. To
have some motivation for the expression of P (B/A ) , consider the case of finite
sample space, where S consists of n elements andA , B have mi and m2 elements
respectively. Let m3 elements be common toA and B. Then referring to Figure 2.3,
P ( B/A ) - Probability of B relative toA
I
no. of elements of B which are also in A
no. of elements in A
m3 m3/n P ( A n B )
- - a-
. ml ml/n - P ( A)
Also
The above discussion suggests the following definition for P ( B/A ).
Deflnition :
Let A and B be events in S and P (A ) 0. Then
Observe that the formula for the conditional probabilityp ( B/A ) given4
equation (2.13) is defined in terms of the unconditional measure P. Therefore,
we have the following two ways of computing the conditional probability
P( B/ A) :
i)
Directly by considering the probability of B with respect to the reduced
sample space S.
ii)
Using the equation (2.13) when P (A n B ) and P ( A ) are computed'with
respect to the original sample space S.
The formula for P ( A/B ) can be obtained from equation (2.13) by
interchangingA and B, i.e.
P( A/ B) = provided P ( B ) L 0.
P ( B )
Example 3 :
If the probability that a research project will be well planned in 0.80 and the
probability that it will be well planned and well executed is 0.72, then find the
probability that a research project which is well planned will also be well
executed.
Solution :
LetA denote the event that the research project is well planned and B the
event, that the project is well executed.
Then P (A ) = 0.80
P ( A f l B) = 0.72.
Required probability is
Example 4 :
Consider a lot consisting of 20 defective and 80 non-defective items. If two
items are chosen at random, without replacement, then find the probability that
both items are defective.
Solution :
Let events A and B be defined as ~ollows:
A = {the first item is defective)
B = {the second item is defective)
ThenP( A) = -
1 19
20 = - and P ( B/A ) - -. Note that here P ( B/A ) has
100 5 99
been computed by evaluating the probability of B in the reduced sample space
A. AfterA has occurred, there are only 99 items in the lot and out of these 19
are defective ( as it is the case of without replacement) and this gives
19
P ( B/A ) - - . Therefore from equation (2.13),
99
2.4.2 Independent Events and Multiplication Rule
Recall the definition of conditional probability given by equation (2.13) and note
that the same can be written as:
P ( A n B ) = P ( A) P(B/ A), P ( A ) * 0 (2.14)
or equivalently
P ( A n B ) = P ( B ) P(A/ B), P ( B ) * 0. (2.15)
These are known as the multiplication rules of probability and are used to compute
the probability of the simultaneous occurrence of two eventsA and B.
If the conditional probabilities are equal to corresponding unconditional
Thwrems of Probabity
probabilities. i.e. P ( A/B ) = P (A ) ( P ( B ) # 0 ) and
P ( B/A ) = P ( B ) ( P (A ) * 0 ), then we are motivated to say that events A and
B are "independent" because then the knowledge of occurrence ofA (respectively B )
in no way influences the probability of the occurrence of B (respectively A).
In view of relations (2.14) and (2.15), provided neither P (A ) nor P ( B ) equals
zero, the relations P ( A/B ) = P (A ) and P ( B/A ) = P ( B ) hold if and only if
The advantage of relation (2.16) is that it is valid even if P (A ) or P ( B ) equals
zero and therefore we take this as the definition of 'independent events'.
Definition :
Any two eventsA and B in a sample space S are said to be independent if
The above definition can be extended for more than 2 events. We say that the
three eventsA , B , C in a sample space S are mutually independent if all the
following
conditions hold:
P ( A n B ) = P ( A ) P ( B )
P ( B ~ C ) = P ( B ) P ( C )
P ( A n C ) = P ( A ) P ( C ) 1 (2.17)
P ( A n B n C ) = P ( A ) P ( B ) P ( C ) (2.18)
If only equations (2.17) hold then we say that A , B and C are pairwise
independent.
You may now try to give this definition for n events and arrive at the follobing :
Definition :
The n eventsAl,Az, ..., A, in a sample space S are said to be mutually
independent if
P( Ai 1 n A; z n ... n A&) = P( Ai 1) P( Ai 2) ... P ( A&)
for K = 2,3, ... n.
Example 5 :
Find the probability of getting four heads in a row with tossings of a balanced
coin.
Solution :
Let Ai denote the event that the i th throw turns up "Head", i = 1,2,3,4.
Assuming that the throws are independent, the eventsAl,A2,A3 and A4 will
he mutually independent and hence the required probability is given by
P ( Ai n Az n A3 n A4 ) = P( A1 ) P ( A ~ ) P ( A3 ) P ( A4 )
Example 6 :
IfA and B are independent events then show t hat z and B are also
independent.
Solution :
The independence ofA and B implies that P ( A n B ) = P ( A ) P ( B ).
Now
c o n s i d e r ~ ( An g ) = P ( A n B )
(by De Morgan's Law)
= 1 - P ( AUB) (by Theorem 2)
= 1 - [ P ( A) + P ( B) - P ( An B) ] (by Theorem 3)
= 1 - [ P ( A) + P ( B) - P ( A) P( B) ]
(as A and B are independent)
= 1- P ( A) - P ( B ) + P ( A) P ( B )
= ( l - P ( A 1) ( l - P ( B) )
which implies that A and Bare independent.
Example 7 :
In certain region the subgrade is predominantly a silty soil with the odd clay
lense. m e average size of these lenses is 750 m2. If on a site 100 x 150 m2 a
clay lense exists, what is the probability of encountering it in any one of the
eight boreholes, symmetrically placed over the site. State assumptions made, if
any.
Solution :
Let F denote the event that the clay lense is found in a borehole. Then, P ( F )
can be estimated from the ratio of two areas: (recall the definition of geometric
probability studied in Unit 1)
Now P ( F ) - 1 - P ( F ) = 0.95 is the probability of not encountering the
lense in one borehole. Hence if the lense does exist, then the chance of
encountering if in atleast one of the eight boreholes is ( F; denoting the event
that it is found in the i th borehole, i = 1,2, ... 8 ).
= 1 - P( Fl UF2 U ... UF8)
(by Theorem 2)
(by De Morgan's law)
<
= 1 - P ( F 1 ) P ( F2 ) ... P ( ) (assuming the independence
F1, .a. F8 )
In solving the problem, we have assumed that the events F'; ( i = 1,2, ... 8 ) are
mutually independent.
Example 8 :
Consider a lot consisting of u) defective and 80 non-defective items. If two items
are chosen at random, with replacement, then find the probability that both items
are defective.
Solution :
Similar to example 4, let eventsA and B be defined as follows:
A = {the first item is defective)
B = {the second item is defective)
Theorems of Probability
20 1
Th e n P( A) = - = - . Now note that, as items are chosen at random with
100 5
replacement, for the second choice there are again 100 items out of which 20
20 -
and P ( B ) = P ( B/ A) , are non-defective. Therefore P ( B ) - - - -
100 5
i.e. eventsA and B are independent..Thus the required probability is
You should read Example 4 again and understand the difference between "with
replacement " and "without replacement" random selections.
A die is tossed, and independently, a card is chosen at random from a regular
deck. What is the probability that
i)
The die shows an even number and the card is from a red suit ?
ii)
The die shows an even number or the card is from a red suit ?
The probability of closing of each relay of the circuit shown in Figure 2.4 is
given by p. If all relays function independently, find the probability that
current exists between the terminals L and R.
H
Figure 2 4
The probability that a worker be late to work on any one day is 0.60. Assuming
that his being late on any one day is independent of his being late on any other
day, find the probability that in a given week he will be late on Monday, on
time on Tuesday, Wednesday, and Thursday, and late again on Friday.
Probability Concepts E8
On a particular building site the lorries delivering materials are fuelled either
by petrol or diesel. It is known that 60% of the lorries deliver sand and 55% of
the lorries use petrol. Determine the probability that a lorry chosen at random
will not be delivering sand and will be diesel powered.
If A and B are independent events then show t hat 2 and B are also
independent.
2.43 Theorem of Total Probability and Bayes Theorem
Ym have already seen the use of the concept of conditional probability to evaluate
the probability of the simultaneous occurrence of two events. Here, you will see that
the concept of conditional probability can be used in another way to compute the
probability of a single eventA when certain "partition" of the sample space S is
given. These results will lead to Bayes Rule or Bayes Theorem for finding the
probability that the "effect7'A was "caused" by the event B. For this, we need the
following definition.
Definition :
The event B1, B2, ... BK are said to define a partition of the sample space S if
( i i ) B; nBj = +,i + j
( iii ) P ( B; ) > 0 for all i .
In the below given figure (Figure 2.5) the events B1, Bz, Bs, B4,85 and Bs
define a partition of the sample space S.
I .
F i i 2 5
It should be clear from the definition that events Bi, B2, ... Bk, define a partition of
the sample space S if the random experiment E results in one and only one of the
events Bi . For example if the random experiment E represents the tossing of a
die, then S = { 1,2,3,4,5,6 } and for this sample space
Bi = [ 1,3,6 } and B2 - { 2,4,5 } define a partition of S. On the other hand
B1 - {1, 4}, B2 = ( 5, 6}, B3 - (2,3,5}donotdefiheapartitionof~
because B2 l l B3 LL (p. We now present the "Theorem of total probabilityw
(Theorem 5):
Theorem 5 :
Let B1, B2, ... Bkbe a partition of the sample space S and letA be some event
with respect to S. Then
Proof:
k
Note thatA - A nS . But s = U Bi, and hence
i - 1
A - An( Bl UB2U ... UBk),i.e.
1
I = U ( AnBi )
i - 1
(The following Venn diagram (Figure 2.6) illustrates this fact for n = 6.)
Now for i @ j, (A n Bi ) and (A l l Bj ) are mutually exclusive as Bi f i Bj = $.
Therefore from the probability axioms, equation (2.19) gives
Theorem6 of Probability
But by the definition of conditional probability P ( A fl Bi ) = P ( Bi ) P (A/Bi )
(recollect the multiplication rule) and hence equation (2.20) implies
Probabfity coIke~~k-3
k
P ( A ) 2 P ( Bi ) P(A/ Bi ),
i - 1
which proves the theorem of total probability.
Example 9 :
An assembly plant receives its voltage regulators from two different suppliers,
75% from supplier B1 and 25% from supplier B2. Suppose that 95% of the
voltage regulators supplied by B1 and 80% of those supplied by B2 perform
according to specifications. What is the probability that a randomly selected
voltage regulator from the assembly plant will perform according to
specification ?
Solution :
LetA be the event that a randomly selected voltage regulator from the
assembly plant performs according to specifications. Note that B1 , B2 define a
, partition of the sample space (as a regulator has certainly to be supplied by
either Bi or B2, but not from both), the Theorem of total probability becomes
applicable. The relevant probabilities can be computed as follows:
Hence by the theorem of total probability,
P ( A ) = P ( B1 ) P ( A/ Bi ) + P ( B2 ) P( A/ B2)
Example 10 :
A building contractor requires a roll of roofing felt. There are 3 suppliers,
B1, B2, B3 in the area and the probabilities (based on his previous experience
and the location of the supplier) that the contractor will instruct his van man to
visit a particular supplier are 0.6,0.2 and 0.2 respectively. Each supplier stocks
roofing felt produced by two manufacturers, X and Y . Both types of roofing
felt sell for the same price and both satisfy the current building regulations.
The stock situation at each of the supplier is:
Supplier
B 1
B2
B3
No. of "X * rolls No. of "Y" rolls
10 30
30 20
30 10
Assuming that the van man will be told by his employer which supplier to
visit, which roll type ( X or Y ) is the van man most likely to return with.
Solution :
Let P ( X) be the probability that the van man will return with roll typeX , P
(Y ") being defined similarly. From the given data,
Hence
p ( X ) = P( B1 ) P( X/ Bl ) +P( B2) P( X/ B2) +P( B3) P(X/B3)
In a similar manner, P ( Y ) - 0.58, Hence it is more likely that the van man
will return with a roll of type Y.
Next we proceed to establish the Bayes Theorem (also called Bayes Rule or
Bayes formula) which helps to evaluate the probability that the "effectmA was
"caused" by the event Bi i.e. P ( Bi/A ) where events Bi, B2, ... Bk represents
the partition of the sample space S.
Theorem 6 :
Let Bi, B2, ... Bk be a partition of the sample space S and letA be some event
with respect to S . Then for 1 s r s k,
I
P mf :
The proof follows immediately from the definition of conditional probability
and the theorem of total probability (Theorem 5) because
Example 11 :
With reference to the data given in Example 9, suppose that a voltage regulator
from the assembly plant is chosen at random and found to perform according
to specifications. Find the probability that it has been supplied by the supplier
B2.
Solution :
LetA, Bi, B2 be the events as defined in the solution of Example 9. So we
1 have
P( B1 ) = 0.75, P( B2 ) = 0.25
and we have to determine P ( BdA ) . This is precisely the situation when
Bayes rule is applicable and hence using Theorem 6,
Probability conceps Emmple 12 :
In a bolt factory, machines Bi, B% and B3 manufacture 25,35 and 40 per cent
of the total output respectively. Of their outputs 5,4 and 2 per cent, respectively
are defective bolts. A bolt is chosen at random and found to be defective. What
is the probability that the bolt come from machine B3 ?
Solution :
Let us introduce the following events:
A I {the item is defective)
B1 = {the bolt came from machine Bi )
B2 = {the bolt came from machine B2 )
B3 - {the bolt came from machine Bg ) .
From the given data, P ( Bl ) = 0.25, P ( B2 ) = 0.35, P ( B3 ) = 0.40,
P ( A/ B1 ) = 0.05, P ( A/ Ba ) 5 0.04, and P ( A/ B3 ) = 0.02. The
required probability is
Remark
The theorem of total probability (Theorem 5) can be visualized by constructing
a tree diagram like that of Figure 2.7, where the probability of the final
outcome is given by the sum of the products of the probabilities corresponding
to each individual branch.
F m 2.7:TrccDiqpm
The tree diagram, so named because of its appearance, can often be useful in
decision making. Each branch illustrate the path that will be taken whenever a
particular decision is made. Figure 2.7 illustrates the tree diagram with
reference to the theorem of total probability only. In general there may be
more than one branch from Bi ( i - 1,2, ... k ) resulting in different events
which may have further branches. The tree diagram for Example 10 is shown
in Figure 2.8 and shows all the possible combinations of events that could be
involved in finishing up with an X or Y type of roofing felt.
Thtomns of Probability
Figure 2.8 : T m Diagram for Exnmple 10
0
Consider a lot consisting of 20 defective and 80 non-defective items. If two
items are chosen at random, without replacement find the probability that the
second chosen item is defective.
There are four boxes numbered 1,2,3 and 4. Box 1 contains 2 red and 8 white
balls, Box 2 contains 5 red and 5 white balls, Box 3 contains 6 red and 4 white
balls and Box 4 contains 9 red and 1 white balb. A box is chosen at random
and then a ball is chosen at random from the box. find the probability that the
chosen ball is red. Use tree diagram to evaluate the required probability.
Consider the data of Example 10 and suppose that the van man returned with
an X type of roofing felt. Determine the probability that he obtained it from
supplier B2.
2.5 SUMMARY
We briefly sum up what has been done in this unit.
1)
Using the axioms of probability, certain results have been derived for
evaluating the probability of combination of events. Specifically we have
proved the following :
i ) P ( $ ) = 0
- i i ) ~ ( x ) = 1 - P ( A)
Suppose that a vacuum tube may come from any one of three manufacturers
with probabilities PI - 0.25, P2 = 0.50 and P3 = 0.25 . The probabilities
that the tube will function during a specified period of time equal 0.1,0.2 and
0.4 respectively for the three manufacturers. Compute the probabiliy that a
randomly chosen tube will work for the specified period of time.
El4
With reference to the data of E 13 above, suppose that the randomly chosen
tube has been found to work for the specified period of time. Find the
probability that it has been manufactured by the first manufacturer.
iv) A C B +- P ( B- A) = P ( B) - P ( A)
V ) A C B +- P ( A ) s P ( B ) .
A general addition rule has been given for finding the probability of
occurrence of at least one of the given n eventsAl,A~, ... An. This rule states
that
P( A1 UA2U ... An) - Z P ( A ~ ) - C P ( Ai n Aj ) + ...... +
i - 1 i c j - 2
The concept of conditional probability has been introduced and P (A/B ) has
been defined as
Using the definition of conditional probability, the multiplication rule for
finding the probability of joint occurrence of two events, has been stated as
P ( A n B ) = P ( A ) P( B/ A) = P ( B ) P( A/ B) .
The concept of mutually independent events has been introduced. Two events
A and B are said to be independent if P (A n B ) = P (A ) P ( B ) . Three
eventsA , B, C are said to be mutually independent if every two of these are
i ndependent andP( AnBnC) = P ( A ) P ( B ) P ( C ) .
The partition of the two sample space S is defined and for a given partition
Bi, B2, ... Bk of S andA any event, theorem of total probability and Bayes
theorem have been proved. These are stated as
and
P( Br ) P( A/ Br )
P( Br / A) respectively
Probability ConccplP
E2
Let A be the event that he will ask to have tyres checked and B be the event
that he will ask to have his oil checked. From the given data,
P ( A ) 5 0. 12, P(B) = 0.29 and P ( A n B ) 5 0.07.
i) Required Probability 5 P ( A U B )
ii) Required Probability = P ( A U B )
E3
P ( A U B U C ) = P(A)+P(B)+P(C)-P(AnB)-P(BnC)
- P ( A n C ) + P ( A n B n C )
Here everything is known except P ( A n B n C ). But
( A n B n C ) c ( An B) a n d h e n c e P( An Bn C) s P ( A n B ) = 0.
This implies that P ( A n B n C ) = 0
Therefore
E4
We know that
P ( A U B ) = P ( A ) + P ( B ) - P ( A n B ) s P ( A ) + P ( B ) ,
as P ( A n B ) r 0. Let us denote B U C by D and therefore
P ( A UB UC ) - P ( A U D ) s P ( A ) + P ( D )
s P ( A ) + P ( B U C )
s P ( A ) + P ( B ) + P ( C )
LetA be the event that the die shows an even number and B be the event that
1 1
the cord is from a red suit. Then P ( A ) = - and P ( B ) - - . Also, under
2 2
the given situation,A and B are independent and so
ii) The required probability is given by P ( A U B ) which equals
3
P ( A ) + P ( B ) - P ( A n B ) i.e. ,.
E6
Let Ei be the event that the i th relay is closed ( i = 1,2,3,4) and E be the event
that there is current between L and R. Then
E = ( E i n E 2 ) U ( E g n E q )
:. P ( E ) = P ( ( E i n E 2 ) U ( E g n E 4 ) ) , a s ( Ei n E2 ) a n d ( Eg n Eq )
are mutually exclusive
= ~ ( ~ l f l ~ ~ ) + P ( E g n E 4 ) - P ( ~ l n E 2 n E s n E 4 )
= P ( E 1 ) P ( E 2 ) + P ( E 3 ) P ( E 4 ) - P ( E 1 ) P ( E 2 ) P ( E 3 ) P ( E 4 )
( as Ej 's are mutually independent )
= p 2 + p 2 - p 4
2 4
= 2 P -P .
E7
As the events are independent, the required probability is given by
P ( late on Monday ) x P ( on time on Tuesday ) x P ( on time on
Wednesday) x P( on time on Thursday ) x P ( late on Friday )
= ( 0.60 ) ( 0.40 ) ( 0.40 ) ( 0.40 ) ( 0.60 )
ES
.. .
I
1 LetA be the event that the lorry delivers sand and B be the event that it is
petrol operated. Then the required probability is
= p ( A n B )
= P ( A ) P ( )
( as A and B are independent ; refer to the
problem given at E 9 )
' E9 I
~ ( A n i i ) = P(A = 1- P( A U B )
#
= 1 - ( P ( A ) + P ( B ) - P ( A n B ) )
= 1 - P ( A ) - P ( B ) + P ( A ) P ( B )
(as A and B are independent)
= QI - P( A) ) ( I - P ( B) ) = P ( A) P( B) ,
which implies that A and are independent:
El0
LetA be the event that thc lifst drawn item is defective and B be the event that
the second drawn item is defective. Then
A ~ A and Aconstitutes a partition of the sample Space S, the theorem of total
probability becomes applicable and we get
El l
Let the events of selecting boxes 1,2,3 and 4 be denoted by
Bi , B2, Bh and Bh respectively.
Then
Let A be the event that the chosen ball is red. Note that Bl, B2 , B3 and B4
define a partition of the sample space and hence
P ( A ) - P ( B 1 ) P ( A / B i ) + P ( B 2 ) P ( A/ B2 )
+ P ( B 3 ) P ( A/ Bg ) + P ( Bq ) P ( A/ B4 )
Hence,
1 2 1 5 1 6 1 9 22
P A ) = ( - 4 ' - ) 10
4 ' 1 0 - ) 4 ' - ) 10
4 ' -1 10 = - 40 = 0.55
The tree diagram for this example can be drawn as follows:
Figure 29 The Tm dia-
The problem is to evaluate the probability of B2 given X i.e. P B i X . By
definition,
( 1
P ( X ) has already been evaluated in Example 10 and is equal to 0.42. From
the given data P ( B2 ) - 0.2 and P ( X/B2 ) - 0.6.
Hence, Theorems of Probability
El3
r
Let Bi ( i=1,2,3 ) be the event that the vacuum tube has come from the i th
manufacturer andA be the event that the tube works for the specified period of
I
time. From the given data:
P ( B 1 ) = 0. 25, P( B2) = 0. 50, P( B3) = 0.25
P ( A/ B1 ) = 0.1, P( A/ B2 ) = 0.2, P( A/ B3 ) - 0.4.
The required probability is
P ( A ) - P ( B 1 ) P ( A / B i ) + P ( B 2 ) P ( A/ B2 ) + P ( Bg ) P( A/ B3 )
- ( 0.25 x 0.1 ) + ( 0.50 x 0.2 ) + ( 0.25 x 0.4 )
= 0.225.
El4
The required probability is P ( Bl / A ) which by Bayes Theorem is given by :