1338 Probability

Chapter

28
Probability
Introduction (5) Mutually exclusive and exhaustive system of events :
Let S be the sample space associated with a random experiment. Let
Numerical study of chances of occurrence of events is dealt in
A1, A2, …..An be subsets of S such that
probability theory.
The theory of probability is applied in many diverse fields and (i) Ai  A j =  for i  j and (ii) A1  A2  ....  An = S
the flexibility of the theory provides approximate tools for so great Then the collection of events A1 , A2 , ....., An is said to form a
a variety of needs.
mutually exclusive and exhaustive system of events.
Definitions of various terms If E1 , E2 , ....., En are elementary events associated with a
(1) Sample space : The set of all possible outcomes of a trial random experiment, then
(random experiment) is called its sample space. It is generally (i) Ei  E j =  for i  j and (ii) E1  E2  ....  En = S
denoted by S and each outcome of the trial is said to be a sample point.
(2) Event : An event is a subset of a sample space. So, the collection of elementary events associated with a
(i) Simple event : An event containing only a single sample random experiment always form a system of mutually exclusive
point is called an elementary or simple event. and exhaustive system of events.
(ii) Compound events : Events obtained by combining In this system, P( A1  A2 .......  An )
together two or more elementary events are known as the = P( A1 ) + P( A2 ) + ..... + P( An ) = 1 .
compound events or decomposable events.
(iii) Equally likely events : Events are equally likely if there Classical definition of probability
is no reason for an event to occur in preference to any other event.
(iv) Mutually exclusive or disjoint events : Events are said If a random experiment results in n mutually exclusive,
to be mutually exclusive or disjoint or incompatible if the occurrence of equally likely and exhaustive outcomes, out of which m are
any one of them prevents the occurrence of all the others. favourable to the occurrence of an event A, then the probability of
(v) Mutually non-exclusive events : The events which are occurrence of A is given by
not mutually exclusive are known as compatible events or m Number of outcomes favourable to A
P ( A) = =
mutually non exclusive events. n Number of total outcomes
(vi) Independent events : Events are said to be independent It is obvious that 0 ≤ m ≤ n. If an event A is certain to happen,
if the happening (or non-happening) of one event is not affected then m = n, thus P(A) = 1.
by the happening (or non-happening) of others.
If A is impossible to happen, then m = 0 and so P(A) = 0. Hence
(vii) Dependent events : Two or more events are said to be
we conclude that 0 ≤ P(A) ≤ 1.
dependent if the happening of one event affects (partially or
totally) other event. Further, if A denotes negative of A i.e. event that A doesn’t
(3) Exhaustive number of cases : The total number of happen, then for above cases m, n; we shall have
possible outcomes of a random experiment in a trial is known as n −m m
the exhaustive number of cases. P( A ) = = 1 − = 1 − P ( A) ,  P ( A ) + P ( A ) = 1 .
n n
(4) Favourable number of cases : The number of cases
Notations : For two events A and B,
favourable to an event in a trial is the total number of elementary
events such that the occurrence of any one of them ensures the (i) A or A or AC stands for the non-occurrence or negation of A.
happening of the event. (ii) A  B stands for the occurrence of at least one of A and B.
 Probability 1339
(iii) A  B stands for the simultaneous occurrence of A and B. (3) When events are independent : If A and B are
(iv) A  B stands for the non-occurrence of both A and B. independent events, then P( A  B) = P( A).P(B)
(v) A  B stands for “the occurrence of A implies occurrence  P( A  B) = P( A) + P(B) − P( A).P(B) .
of B”. (4) Some other theorems
Problems based on combination and permutation (i) Let A and B be two events associated with a random
experiment, then
(1) Problems based on combination or selection : To solve
(a) P( A  B) = P(B) − P( A  B)
n!
such kind of problems, we use n Cr = . (b) P( A  B ) = P( A) − P( A  B)
r!(n − r)!
(2) Problems based on permutation or arrangement : To If B  A, then
n! (a) P( A  B ) = P( A) − P(B) (b) P(B)  P( A)
solve such kind of problems, we use n Pr = .
(n − r)! Similarly if A  B, then

Odds in favour and odds against an event (a) ( A  B) = P(B) − P( A) (b) P( A)  P(B)
• Probability of occurrence of neither A nor B is
As a result of an experiment if “a” of the outcomes are
favourable to an event E and “b” of the outcomes are against it, P( A  B ) = P( A  B) = 1 − P( A  B)
then we say that odds are a to b in favour of E or odds are b to a (ii) Generalization of the addition theorem : If
against E. A1 , A2 ,....., An are n events associated with a random experiment,
Thus odds in favour of an event E
 n  n n n

=
Number of favourable cases a a /(a + b) P(E)
= = = .
then P  Ai  =
    P( A ) −  P ( A  A ) +  P( A  A  A ) +
i i j i j k
Number of unfavourab le cases b b /(a + b) P(E )  i=1  i=1 i, j =1 i, j, k =1
i j i j  k
Similarly, odds against an event E
... + (−1)n −1 P( A1  A2  .....  An ) .
Number of unfavourab le cases b P(E )
= = = . If all the events Ai(i = 1, 2..., n) are mutually exclusive, then
Number of favourable cases a P(E)
 n  n

Addition theorems on probability P  Ai  =

    P( Ai )
 i=1  i=1
Notations : (i) P( A + B) or P( A  B) = Probability of i.e. P( A1  A2  ....  An ) = P( A1 ) + P( A2 ) + .... + P( An ) .
happening of A or B
(iii) Booley’s inequality : If A1 , A2 , .... An are n events
= Probability of happening of the events A or B or both
associated with a random experiment, then
= Probability of occurrence of at least one event A or B
(ii) P(AB) or P(AB) = Probability of happening of events A  n  n  n  n

and B together.
(a) P 
  Ai  
 
P( Ai ) − (n − 1) (b) P Ai  
  
P( Ai ) 
 i=1  i=1  i=1  i=1
(1) When events are not mutually exclusive : If A and B are
These results can be easily established by using the Principle
two events which are not mutually exclusive, then
of mathematical induction.
P ( A  B) = P ( A ) + P ( B ) − P ( A  B)
or P( A + B) = P( A) + P(B) − P( AB) Conditional probability
For any three events A, B, C Let A and B be two events associated with a random
experiment. Then, the probability of occurrence of A under the
P( A  B  C ) = P( A) + P(B) + P(C ) − P( A  B)
condition that B has already occurred and P(B)  0, is called the
− P(B  C ) − P(C  A) + P( A  B  C) conditional probability and it is denoted by P(A/B).
or P( A + B + C) = P( A) + P(B) + P(C) − P( AB) − P(BC ) Thus, P(A/B) = Probability of occurrence of A, given that B has
already happened.
− P(CA ) + P( ABC )
P( A  B) n( A  B)
(2) When events are mutually exclusive : If A and B are = = .
P(B) n(B)
mutually exclusive events, then n( A  B) = 0  P( A  B) = 0
Similarly, P(B/A) = Probability of occurrence of B, given that A
 P ( A  B ) = P ( A) + P ( B ) . has already happened.
For any three events A, B, C which are mutually exclusive, P( A  B) n( A  B)
P( A  B) = P(B  C) = P(C  A) = P( A  B  C) = 0 = = .
P( A) n( A)
 P( A  B  C ) = P( A) + P(B) + P(C ) . • Sometimes, P(A/B) is also used to denote the probability of
The probability of happening of any one of several mutually occurrence of A when B occurs. Similarly, P(B/A) is used to denote
exclusive events is equal to the sum of their probabilities, i.e. if the probability of occurrence of B when A occurs.
A1 , A2 ..... An are mutually exclusive events, then (1) Multiplication theorems on probability
P( A1 + A2 + ... + An ) = P( A1 ) + P( A2 ) + ..... + P( An ) (i) If A and B are two events associated with a random
experiment, then P( A  B) = P( A) . P(B / A) , if P(A)  0 or
i.e. P( A ) =  P( A ) .
i i P( A  B) = P(B) . P( A / B) , if P(B)  0.
 1340 Probability
(ii) Extension of multiplication theorem : If A1 , A2 , ...., An outcome of an experiment. The probabilities P(Ei), i = 1, 2, ….., n
are n events related to a random experiment, then are called prior probabilities. Suppose the experiment results in
an outcome of event A, where P(A) > 0. We have to find the
P( A1  A2  A3  ....  An ) = P( A1 )P( A2 / A1 )P( A3 / A1  A2 ) probability that the observed event A was due to cause Ei, that is,
....P( An / A1  A2  ...  An −1 ) , we seek the conditional probability P(Ei / A) . These probabilities
where P( Ai / A1  A2  ...  Ai−1 ) represents the conditional are called posterior probabilities, given by Baye’s rule as
P(Ei ).P( A / Ei )
probability of the event Ai , given that the events A1 , A2 , ....., Ai−1 P(Ei / A) = n .
have already happened.  P( E ) P( A / E )
k k
(iii) Multiplication theorems for independent events : If A k =1

and B are independent events associated with a random

experiment, then P( A  B) = P( A) . P(B) i.e., the probability of Binomial distribution
simultaneous occurrence of two independent events is equal to (1) Geometrical method for probability : When the number
the product of their probabilities. By multiplication theorem, we of points in the sample space is infinite, it becomes difficult to
have P( A  B) = P( A) . P(B / A) . Since A and B are independent apply classical definition of probability. For instance, if we are
events, therefore P(B / A) = P(B) . Hence, P( A  B) = P( A) . P(B) . interested to find the probability that a point selected at random
from the interval [1, 6] lies either in the interval [1, 2] or [5, 6], we
(iv) Extension of multiplication theorem for independent cannot apply the classical definition of probability. In this case we
events : If A1 , A2 , ...., An are independent events associated with a define the probability as follows:
random experiment, then Measure of region A
P{x  A} =
P( A1  A2  A3  ...  An ) = P( A1 )P( A2 )...P( An ) . Measure of the sample space S
By multiplication theorem, we have where measure stands for length, area or volume depending
P( A1  A2  A3  ...  An ) = P( A1 )P( A2 / A1 )P( A3 / A1  A2 ) upon whether S is a one-dimensional, two-dimensional or three-
dimensional region.
...P( An / A1  A2  ...  An −1 ) (2) Probability distribution : Let S be a sample space. A
Since A1 , A2 , ...., An −1 , An are independent events, therefore random variable X is a function from the set S to R, the set of real
numbers.
P( A2 / A1 ) = P( A2 ), P( A3 / A1  A2 ) = P( A3 ), ...., For example, the sample space for a throw of a pair of dice is
P( An / A1  A2  ...  An −1 ) = P( An ) {11, 12,  , 16
Hence, P( A1  A2  ...  An ) = P( A1 )P( A2 )....P( An ) . 21, 22,  , 26
S =
(2) Probability of at least one of the n independent events    
: If p1 , p 2 , p3 , ........, pn be the probabilities of happening of n 61, 62,  , 66 }
independent events A1 , A2 , A3 , ........, An respectively, then Let X be the sum of numbers on the dice. Then
X (12) = 3, X (43 ) = 7 , etc. Also, {X = 7} is the event {61, 52, 43, 34,
(i) Probability of happening none of them
25, 16}. In general, if X is a random variable defined on the
= P( A1  A2  A3 ......  An ) = P( A1 ).P( A2 ).P( A3 ).....P( An ) .
sample space S and r is a real number, then {X = r} is an event.
= (1 − p1 )(1 − p2 )(1 − p3 )....(1 − pn ) If the random variable X takes n distinct values x1 , x 2 , ...., x n , then
(ii) Probability of happening at least one of them {X = x1 } , {X = x 2 }, ...., {X = x n } are mutually exclusive and
= P( A1  A2  A3 ....  An ) = 1 − P( A1 )P( A2 )P( A3 )....P( An ) . exhaustive events.
X = x1 X = x3
= 1 − (1 − p1 )(1 − p2 )(1 − p3 )...(1 − pn )
X = x2
(iii) Probability of happening of first event and not happening
of the remaining = P( A1 )P( A2 )P( A3 ).....P( An )
X = x4
= p1 (1 − p 2 )(1 − p3 ).......(1 − pn ) X = xn

Now, since (X = x i ) is an event, we can talk of P(X = x i ) . If

Total probability and Baye’s rule
P(X = x i ) = Pi (1  i  n) , then the system of numbers.
(1) The law of total probability : Let S be the sample space
 x1 x 2  x n 
and let E1 , E2 , .....En be n mutually exclusive and exhaustive   is said to be the probability distribution
events associated with a random experiment. If A is any event  p1 p 2  pn 
which occurs with E1 or E2 or …or En, then of the random variable X.
The expectation (mean) of the random variable X is defined as
P( A) = P(E1 ) P( A / E1 ) + P(E2 ) P( A / E2 ) + ... + P(En ) P( A / En ) . n

(2) Baye’s rule : Let S be a sample space and E1 , E2 , .....En be E( X ) = p x

i =1
i i and the variance of X is defined as
n
n mutually exclusive events such that E = S
i=1
i and P(Ei )  0 for
var( X ) = 
n
pi (x i − E( X ))2 =
n

px 2
i i − (E( X ))2 .
i =1 i =1
i = 1, 2, ……, n. We can think of (Ei’s as the causes that lead to the
 Probability 1341
(3) Binomial probability distribution : A random variable X
which takes values 0, 1, 2, …, n is said to follow binomial
distribution if its probability distribution function is given by
P(X = r) = n Cr p r q n −r , r = 0, 1, 2, ....., n
where p, q > 0 such that p + q = 1.
The notation X ~ B(n, p) is generally used to denote that the
random variable X follows binomial distribution with parameters  Independent events are always taken from different
n and p. experiments, while mutually exclusive events are taken from a
single experiment.
We have P( X = 0) + P( X = 1) + ... + P( X = n) .  Independent events can happen together while mutually
exclusive events cannot happen together.
= nC0 p 0 q n −0 + n C1 p1q n −1 + ... + nCn p nq n −n = (q + p)n = 1n = 1
 Independent events are connected by the word “and” but
Now probability of mutually exclusive events are connected by the word “or”.
(a) Occurrence of the event exactly r times  Number of exhaustive cases of tossing n coins
simultaneously (or of tossing a coin n times) = 2n.
P( X = r) = n Cr q n −r p r .  Number of exhaustive cases of throwing n dice
simultaneously (or throwing one dice n times) = 6n.
(b) Occurrence of the event at least r times
 Probability regarding n letters and their envelopes : If
n
n letters corresponding to n envelopes are placed in the
P( X  r) = n Cr q n −r p r + ... + p n = 
X =r
n
CX p X qn− X . envelopes at random, then
(i) Probability that all letters are in right envelopes = 1 / n ! .
(c) Occurrence of the event at the most r times 1
r (ii) Probability that all letters are not in right envelopes = 1 −
.
p
n!
P(0  X  r) = q n + n C1q n −1 p + ... + n Cr q n −r p r = X n− X
q .
X =0 (iii) Probability that no letter is in right envelopes
1 1 1 1
• If the probability of happening of an event in one trial be p, = − + − ... + (−1)n .
then the probability of successive happening of that event in r 2! 3! 4 ! n!
trials is p r . (iv) Probability that exactly r letters are in right envelopes
1 1 1 1 1 
• If n trials constitute an experiment and the experiment is =  − + − ..... + (−1)n −r .
r!  2! 3! 4 ! (n − r)!
repeated N times, then the frequencies of 0, 1, 2, …, n successes are
given by N .P(X = 0), N .P(X = 1), N .P(X = 2), ...., N .P(X = n) .  If odds in favour of an event are a : b, then the probability
a
(i) Mean and variance of the binomial distribution of the occurrence of that event is and the probability of
a+b
The binomial probability distribution is b
non-occurrence of that event is .
X 0 1 2 ..... n a+b
P( X ) n n
C0 q p 0 n
C1q n −1
p n
C2q n−2
p ..... Cn q p n
2 n 0  If odds against an event are a : b, then the probability of the
b
The mean of this distribution is occurrence of that event is and the probability of non-
a+b
n n

 X p =  X.
a
i i
n
C X q n − X p X = np , occurrence of that event is .
i=1 X =1 a+b
 Let A, B, and C are three arbitrary events. Then
The variance of the Binomial distribution is  2 = npq and
Verbal description of event Equivalent set theoretic notation
the standard deviation is  = (npq ) . (i) Only A occurs (i) A  B  C
(ii) Use of multinomial expansion : If a die has m faces (ii) Both A and B, but not C occur (ii) A  B  C
marked with the numbers 1, 2, 3, ….m and if such n dice are (iii) All the three events (iii) A  B  C
thrown, then the probability that the sum of the numbers occur
exhibited on the upper faces equal to p is given by the coefficient (iv) At least one occurs (iv) A  B  C
(x + x 2 + x 3 + .... + x m )n (v) At least two occur (v) ( A  B)  (B  C)  ( A  C)
of x p in the expansion of .
mn (vi) One and no more occurs (vi) ( A  B  C )  ( A  B  C )
(4) The poisson distribution : Let X be a discrete random  ( A  B  C)  ( A  B  C)
variable which can take on the values 0, 1, 2,... such that the
(vii) Exactly two of A, B and (vii) ( A  B  C )  ( A  B  C )
probability function of X is given by
C occur
 ( A  B  C)  ( A  B  C)
 x e −
f ( x ) = P( X = x ) = , x = 0,1,2,....
x! (viii) None occurs (viii) A  B  C = A  B  C
where  is a given positive constant. This distribution is (ix) Not more than two (ix) ( A  B)  (B  C)  ( A  C)
occur −( A  B  C )
called the Poisson distribution and a random variable having this
distribution is said to be Poisson distributed. (x) Exactly one of A and B (x) ( A  B )  ( A  B)
occurs
1342 Probability