I Solved Problem 1.

7 For a certain binary communication channel, the probability that a

transmitted 'O' is received as a ' O' is .95 and the probability that a transmitted '1' is re-
ceived as T is 0.90. If the probability that a ' O' is transmitted is 0.4, find the probabilit:·
that
(a) a ' l ' is received (ii) a ' l ' was transmitted given that
a ' 1' was received

Solution:
Let

A = event that ' 11was received

B1 = event that ' l'was transmitted
1
B2 = B 1 = event that '0 was transmitted

Then to find P(A) and P (B i/ A)

(a) P (A) = P (A/ B 1) P (B 1) + P (A / B 2 )P (B 2 )
= 0.90 X 0.6 + (1 - 0.95) X 0.4 = 0.56
b) P(Bif A) = P (A / B1 )P (B 1) _ .90 x .6
~ (A/ B1)P(B1) + P(A/ B 2 )P(B 2 ) - 0.56
2
28

Random Variables
Definition: \
 Pn)/1;1hi/it, ;111d N11111/lt·r J'/1cn1'
1.20
. Ii' w l H'll' S' ts lilt; sa111ple spnt.:e ·iss .
~1 10
. ) is ., lurn·t1on I llHI 1 I , . 0(;1a11:d ' ,
.\ 1;1nth)1\t , :H·1:1hk
1
· • (t. \ ., . sl'I ul ,e: tl 1111111lw1 s.
IS ( 1ll '
1
' . l , )l'I iilll'II( :ltld /t r '
,,1tha1,\\Hl)I\\ L\l . ,.· ,(wocoi11s om·l·. llie11
1 ' rtllll'lll ol (OSS III ~~
F"mnpk: ('\,11sidl't t ltl' l'\I \ ,
-- I') (t II) (1 1, I)} .
~ = HJI, //), tll, . 2 \ (II I ') = .\('! ', II ) = I,
~
I I -

l ' \ -4 u b1.' 1.klit11.'d ;\ S \ (II, I I ) - I ,

l.l • . . 1 r·, ·t \ tknotcs the number ol ltemls.

\ ~ f, J ) - (), rlh'll \ tS ,\ l \ . l\ •L

Events defined by R.Vs:-

'1.' ,l I'.\'. ,rnd .r lk' a (1\.L'd real number. Then Lhe uvcnt f.X ~ :c] is defined as
Ll't .\

[X .r] =- {w E S : X (w) == x} .

S,milarl\. [.\" -:-- .r] = {w ES : X(w) < x} and

l(: , X , b] = {w E S : a < X (w) < b} .
ln the abo, e example [X = O] = {(T, T)} .

Distribution function (df) (or) Cumulative distribution

function (cdf)

Let X he a r.v. Then the cdf of X, denoted by .Fx(x), is defined as

I .Fx (x) = P [X ~ x] \

the probability of the event [X < x] .

Properties of cdf:

L0:5Fx(x):5 1

2. Fx (-oc) == o
l Fx (oc) =l
~ al = Fx (b) -
4
· p [a .,,, X :5 b) = P [X : b] P [.X F, (a)

5. p [X > a] =1 - I' [X :S ,, l == 1 p\ ( o)

Discrete random variable:

A r.v. Xis said to be a discrete , ·1· v· k . . .. " valttt'~.

"' I , V I ,.(\ , tn res on ,l liuile or cm111tahl) ll\tillth:

j
 ·i·ty Mass Function (pmf):
probab11
. ete r ·v X is denoted by p -x ( x) and it is defined as
Thep mf 0 fa d1scr
/~ P-
p j-x (-x)-=- ]/
[1Y_ =_ x---.

Properties of pmf:
I. Px(x) > 0, Vx
2. ~Px(x) =1
X

3. Fx (a) = P [X < a] = L P [X = x] = L Px (x) (expression for the cdf in

x~a x ~ a
terms of pmf)

Expectation:

If Xis ar.v., then the expectation of X (or Mean), denoted by E [X ], is defined by

Mean = E [X ] = L XiPx (x i)
'l

2] = L Xf Px (xi)
1. E [X
i
2
- Variance of X = Var(X) = E [X 2 ] - {E [X]} 2

Problems on Discrete Random Variable:

1· A r.v. X has the prnf

X -2 -1 0 1 2 3
p(x) 0.1 K 0. 2 2K 0.3 3K
(a) p·
Ind the value of K
(b) Evaluate P (X ·
(c . < 2) and P(- 2 < .rY < 2).
J Find the cdf of X
(d) Fi d .
n the rnean of X.
Solution:
(ci Cdf of X: X
I I ll l!
ll -
-00 -2 -1 0 2 3 co
(i) Let -co < x <-2
Then,
F(x) =0 X

-00 -2 -1 0 1 '2 3 co
(ii) Let-2 ~ X < -1
Then, 1
F(x) = P[X ~ x) = P(X = -2) = 0.1 = lO
X

(iii) Let-1 ~ x < 0. -2 -1 0 1 2 3 rti

Then,

F(x) = P[x < x] = P[x == -21+ P(x == -1)

1 1 5 1
- 0.1 , k == - + - = _ == _
_ I

10 15 30 6
X
(iv) Let Os; X <1
Then, -2 -1 0 ; 2 3 1.
 Proba/)1/ity ancl p , 1
\, 1t1< om Variables
1.23

P(.r) == Pl \ ,., :r;j /J ( X ~ I]

= (P[ \ = - 2] -1- JJ[ X = - l ) f P [X = O]) P,X = J)
1 l. 2 15 1
= 30 + 15 - 30 = 2
X
-oo -1 0 1 2
-2 3 'l,

F(x) = P[X ~ x] = P(X < 2]

= (P[X = -2] + P[X = -1] + P [X = O)]
+ P[X = l]) + P [X = 2]
1 1 3 8 4
= - + 0.3 = - +- = - = -
5
2 2 10 10

-2 -oo -1 0 2 3 co
(vii) Let 3 ~ x < oo.
Then,
F(x) = P[X < x] = P[X < 3] = Total probab ility= 1
Thus the cdf is given by

0 -00 < X < -2

'
1
- -2 < x < -l
10 ' -
1
- -l<x <O
6' -
11
F(X) = -,O< x<l
30 -
1
- l <x< 2
2' -
4
-,2< x<3
5 -
1, 3 <:r<c x.:
 \<l) i-,li:-an of X

-
_ ,) ( 1 )
IO
5
,
I( i )
- 3_ l +2+9+9
( :5) +
( 10) + ( { )
+0 + l
2 3

16
1 I 2 3 3 _ ____::_---:-----=-
'J, +- +- +- -
- - 15 15 5 5
== -)05
15 15

bTt · distribution:
2. Ar., X l1:1s the followmg proba 11 y

3 4 5 6 7
X 0 l 2
p(x) 0 k 2k 2k 3k k2 2k 2 7kL. +k

Find (i) the value of k, (ii) P{l.5 < X < 4.5/ X > 2} and (iii) the smallest Value of
1
)dor which P(X ~ J\) > 2°

Solution:

(i) Total probability is =1

⇒ 0 + k + 2k + 2k + 3k + k 2 + 2k 2 + 7k2 + J..: =l
2
⇒ 10k + 9k - l = 0 ⇒
10k 2 + 10k - "It: - 1=O
⇒ lOk(k+l)-l(k+l)=O

⇒ (lOk - l)(k + 1) == 0 => lOk - 1 == 0 ⇒ 10k == 1 ~ k == ]:_

10

(ii) p [ (1.5 ; ~; 4.5)] == P[(l.5 < X < 4.5) 'l (.x· > 2))
P (X > 2)
== P(X == 3) + P(X == 4)
1 - P( X < 2) - - p (_y == 3) + P IX - 4)
1
:::: _ 2k + 3k Sk [~{-Y == 0) + P (~Y == 1; + P ' X == 2)]
1 - [O + k + 2k] == - - === "/io 5 '10 .J
1 - 3k 1 - 3/10 - ', 10 == -;
 Prohab1/11y and H·, I
, nc o,n Variables
1.25
1
~ < 01 :::: 0 -:f 2
•'') J'[.,\
l111 ~

[ 1 \ : 1] == P[ \ = 01·\ 1l~{ l o+ k ::.::

=== 1 === 1 f ~
~ I() 2
J1 \ ; 2] .; ( P[ \ - O] + P( Y - I) + fJ[ X = 2j
I 1. l 2 :3
2
= 10 \- k = 10 + 10 == 10 f ½
Pl-\ ::, 3] = (P[X = O] + P[X = l] + P[X = 2)) + P[X = '.lJ
= -3 + 2k = -3 + -2 == -5 == _1 -:f _1
10 10 10 10 2 2
1
P[X < 4] = - + P[X = 4] = -12 + 3k = -1 + -3 8
== _ == 8 > ~
- 4 2 10 10 . 2

Therefore' the desired value of A is 4

,., ~e probability function of an infinite discrete distribution is given by

:/PX==J.] -_ 2_., 1· = 1, 2, 3, .... Verify that the total probability is 1 and find the
mean and van·!ce of the distribution. Find also P (X is even), P (X 5) and P (X is ~
divisible by 3).
Solution:

P[X = j] = ~ ,j = 1, 2, 3, ...
X 1 2 3 n

I P(x)
1 12 1
- - -
3

2 2 2
Total Probability is

1 (1)2 (1)3
= 2 + 2 + 2 + ... -
- 1 -1/ 1/2
2 = 1/2 = 1
1/2

Mean of X = E[X] = ~n Gr= G) (ff +2 +3 (ff +· ..

2
== x + 2x + 3x3 + .... , where x = 1/2
== x(l + 2x + 3x 2 + .... )
X 1/2 - 1/ 2 == 2
= x(l - x)-2 = (1- x)2 = (1 - 1/2)2 - (1/2)2
 n, ~,n
, 6 \
'
' l ,.,,
\

\ --" t ) 1) n] (112t

Therefore,
2 2
Var(X) (E(X)] = 6 - 2 = 6 - 4 = 2
= E (X 2) -
P(X is even) = P (X = 2) + P (X = 4) + P(X = 6) + ...

=Gr +Gr +G)6+ ... 2

(½) ¼ 1/4 1
= 1 - (½) 2 = 1 -1/4 = 3/4 =3
P(X ~ 5) = P (X = 5) + P(X = 6) + ...

=Gr+ Gr+ Gr+ ... .

5
(½) 1/32 1
= --1 = - = -
1- 2 1/2 16
P!X is divisible by 3) = P (X = 3) + P(X = 6) + P(X = 9) + ....

=G)3+Gf +G)9+ .. . -,•:~\·,. ;

(1)3
- l 1
\p'--

= 2 3 =-L-8 1
1- (½) 1-½-i=7
Note: rt t-w· + u,,.2 + _ a.. ·r
... - y::- 1 - 1 < T < l
r ..
4. ff the r.v. X takes ihc v·1lu I21
/\P [X 3) = 5J' [X _ ] ' . es ' ' and 4 such that 2P [~\"" -= L) = 3P (.Y == 2]::::
' - '1 , ftncl the pmf and cdr or V
Solution : · \. ·

2/'IX -= I] :~P[ X 'lj (1)

P[ X = 31 = 5P[_Y = 4]
 l'roh.i/J1/11 y <.11Jd Ua11dom V: .
c1nab/e':, 1.27

, . 1 ,rob:\bt l,t, .. 1
1\1ta l :::;- .P[ \ - t] + P[X = 2] + P[X = :3] + P [X = 4) -= 1
2.1· 2x
_,__
-, .. + -3 + 2X + -5 = 1
l'

l5x + l0x + 30x + 6x 1 15

==}
15
= * 6lx = 15 ⇒ x =
61
2x 10 30 2x 6
2
==} 3 = 61' x = 61 ' 5 = 61
Therefore, the prof of X is
X 1 2 3 4
15 10 30 6
P[X = x} 61 61 61 61
cdf of X:
X

2 3 4 co
(i) Let -oo <x <1
Then,
F(x) = P[X < x] = 0
X

(ii) Let 1 ::; x < 2, -co 2 3 4 oc

Then,
15
F(x) = P[X < x] = P[X = 1) =
61
X

(iii) Let 2 < x

-
< 3' -co 1 2 3 4 :x:

Then,

F(x) = P[X < x] = P[X = 1) + P[X = 2)

15 10 25
= 61 + 61 = 61
X

Civ) Let 3 ~ x < 4. -co ·1 2 3

4 ::X,,'

F ( X) == p [X < X1= p [X = 1] + p [X = 2] + P[~\ = 3]

25 30 55
= 61 + 61 = 61
 Ptobab1lity and Numh~r 777cilf.)' 2 3

ll'lt'•n. . -= 2] + P[X = 3] + P[X ::::: 4]

, 1] + P[;-\
F(x) ; l \ : :z:J == er\
=l
Thu'-. th~ cdf i~ g n ~n b} o, -00 < X < l Pre

~ 1 <x <2 1.
61 ' -

F (x)
~ 2 <x< 3
=== 61 ' -
55
2< X <4
61 ' -
1, 4 < X < 00

Continuous Random variables:

A r.v. X is said to be a continuous r.v, if there exists a non-negative function f x (x), where
x E (-oo, oo), having the property that for any set A of real numbers, p (XE A) ::
J
A
fy (x ) dx.

The function f x (x) is called the probability density function (pd:f) of the r.v. X and 6
defined by
d
fx (x ) = dx Fx (x)

Properties of fx(x):

I. fx (x) > 0, 'vx

00

2- j f x(x)dx = l
- oo
b
3
· P(a ,:_ X ~ b) = .f f x(x)dx = F (b) - F (a)
a

Jf ( )
a,

4. P( X - a) -- XX dX =Q
a
 Probability and Random Variables 1.29

5. P( ,\ < a) == P(X < a) = P( -oo < X < a) = j f x(x )dx

-oo

Problems on continuous Random Variable:

:i:2
xe-2 , x > 0
1. If j(x~ == { 0 ' X<0

(a) show that J(x) is a pdf.

(b) find the cdf.

Solution:
212 x2
Consider: / xe-x dx Putt= -
2

-t
ix
=
=
J t
e- dt= -

-e-x2 ; 2
-1
e
dt = -dx
7
(1)

x<O x~ O
(a) (i) Clearly f (x) > 0, f(x)=O 0 f(x)=xe·.'12 co
for all x .
00 0 00

(ii) j f(x)dx = j f(x)dx + j f (x )dx

-oo -oo O

00

= j xe-x /2dx = [-e-x'/2]:,

2

0
= -[0 - e0] = 1

Hence, f (x) is a pdf.

X
(b) (i) Let -oo < x < 0. Then, -00 0
F(x) = P[X < x] = O
 t
,J ,\ ,,mh'' I /u•otJ' .t
Pr(Jbabi1i '
(1110
1.30 0

.
01) tct O' .r ,I'

f (:l') d:r:
l'[X < .1:) - ./
F(.d N
,1:

_J 0
r (O) / f (r)dJ:
f{<t:),Lr 1· O · ·

- 'v ] X
x
[e -x 2 /2 , by(l)
=
J xe -x2;2dx == - O

= -O [e-x2 /2 - e° ] = - [e-x /2 _
2
1] = (1 - e-x /2) 2

[~us,
Q if - 00 < <0 X

F(x) = { 1- ex2;2 if O < x < 00

2. If the density function of a continuous r. v X is given by

(ii
O< x <1
Il
ax,
a, l<x<2
f (x) = 3a - ax, 2 < x < 3
O , elsewhere
i. :find the value of a,
ii . find the cdf of X and

iii. if x1, x2, x3 are 3 independent observations of X, what is the probability thj
exactly one of these 3 is greater than 1.5?
Solution:

X
(i) f (x) is a pdf f(x)==O
I I I

0 1 2 3
~ 3
f(x)=ax f(x)=a f(x)- a~a,

00

'=> j f(x)dx == 1
- 00
 z iooam my"nna Random Variables 1.31
2 1

j J(x)dx + .I f(,·)d.r 2f f(:r)dx

I

J
0

,=;. f( :i:)dx + --1

-C() 0 I
"<)

+ j f(x)d. r =1
3
2 3

,=;. l axdx + j adx + j

1 2
(3a - ax)dx =l
1 3
⇒
2 2
a(x + a( x) + (3ax - 2 ax ) =1

n-
)
2 0 1 2 2

,=;. a ( D+ a(2 - 1) + [3a(3) - a (

3
[3a(2) - a (~
2
1= l
a 9a
⇒ - + a + 9a - - - 6a + 2a = l
2 2
= 1⇒ Ja = 1⇒ a= ~
2
25a _ 21a
⇒ 2 2 2. i
(ii) cdf of X:
(i) Let -oo < x < 0.
X

- 00 0 1 2 3
Then,
X

F (x) = P[X < x] = j f(x)d x = 0

-CX)

(ii) Let O < x < l.

X

00
-oo 0 1 2 3
Then,
X

F(x) = P[X < x] = j j(x)d x

-CX)

0 (O)
j fet}dx + j j (x)dx
X

=
-CX) 0

x2 1 ( 2x2) = 4 2

i x ( x2 ) x x
= axdx = a =a = 2 2
2 0
0

(iii) Let 1 < x < 2.

1.32
-----· ."t"

-~~ --;-~
~

'fh~ll.
r'

_ ; · /(.r)d.r
F(,l)
~ P[.1; ::f J;] -
- cv X

~ ~(O) )1. J(x)dx + j

= J
- c,O
j(.t)d.1: + 0 l
f (.r)dx

j
X

= axdx + j adx
0 l

_- a (~ 2) 1+ a(x)f
2 o

=a(½) +a(x-l)

(1)
1
1
=2 2 +2x -( _ 1)
11 1
= -4 + -2 x - -2
1 1
= -x--
2 4

(iv) Let 2 :S x < 3.

:c
Then,
0 1 2 3

F(x) = P[X < x] = j f(x)dx

-oo
(0) 1 2

J
Q X

=
- oo
JJ:tfdx+
O
j f (x) dx + j f(x)dx + j f (x)dx
1 2
1 ')

=/ axdx + j +] adx (3a - ax)dx

I 2
~ ( X2) 1+ a(x)
a ...:_
2
2
1 + ( 3ax - a
.-,
x-
)
T

0
2 2
 Pm/JahilJty :m<l f~andom Variahles 1.33

I •, tlo,
_ (1, j- (1 j- ,JO,:t
..
•) 2
n.r 2 fJ
::::: :~air; - - n.
2 2
:3 x2 5
- -:.r - - - -
- 2' 4 4

(,1) I d 3 :::-- ,r < 00 · .x

-oo 0 1 2 3
Then,
X

F(x) = P[X < x] = j f(x)dx

-00

o (0) 1 2 3 oo (0)
= j fet)dx + j f(x)dx +j f(x)dx +j f(x)dx + j fet)dx
-00 0 1 2 3
0 2 3

= j axdx+ j adx+ j(3a-ax)dx

-00 1 2

= a ( x: ) : + a (x) i + ( 3ax - ax: ) :

=;+a + (9a -
9
n- (6a - 2a)

=!a+ a+ ~a -
2 2
4a = 6a - 4a = 2a = 2 (!)2 = 1

Thus:
0, -00 < X <0
x2
, 0 <x<l
4
1 1
F(x) = -X -
2 4' 1 -<
- X < ')...,
2
3:r 1: 5 •)
- - - -, 2 < :l~ <.... v
2 4 4 -
1, 3 < .r <oo
- - - ------::;----=-:- - --:--:-:-;-:-:--::;:~d4 N11I11hC'l 1 11,•pr;
1.34 J>rob:1lnl1lJ an

(iii) I'[~\ '> 1.5, 1 >[1.5 <.. . \ ~ < c,,J

- /<".,{N) r,'(1.G )I [ I (;3)
I ~
[ ( I. 5) - ,I ] =I- 2 ?. . ¾]
:l
] - [ ,J
-
I
-t
J-- l . l -
2
L
2

. onl,;._ 0 f the observations greater than l.5 is given by

Therefore. th1.' P•1.,t,,111ilav th,ll
1
p=-
2
H~nc\.:\ the probab1hty
• that one o f the observations less than 1.5 is given by
1 1
q=l-p=l-2=2

Therefore. out of 3 observations exactly one of them is greater than 1.5 is

= pqq + qpq + qqp

= 3pq2

= 3
G) Gr 3
-
8
3. The cdf of a continuous r. v. X is given by

0 x<O
x2 1
, O<x<
F(x) 2
3 2 1
1- (3 - x) , < X < 3
25 2
1 , X >3

Find the pdf of X, evaluate P (IX/ < I) and P (~ < X < 4) using both pdf and cdf
Solution:
Cdf is given by
O,x < 0
2
x,O<x<½
F(.c) =
l - ~
25
(3 - x) 2 ~<x<3
'2 -
l, ..c > 3
 Prob:i/Jil ity a nd Un ndom ½1riahlcs 1.35

. js criYl'll by
.. , 1hc PJI ' t;-,
ThcrclcHl . 0, :r, < ()

df(.r)
2~r., () : .r : I h.
f (.r,) = d:r fj I
2fj (:~ - :r;)' 2~ X < ~J
0, :r, > ;3

U~ing pdf:
P[j \ j ~ l] = P[- 1 < X :s; 1]
1

= j J(x)dx
- 1

100 25
4

PG~ X<4) = j f( x)dx

1/3
1/2 3 4 ( )

= j f (x) dx + j f (x)dx + j Jk-t'Jd,x0

1/3 1/ 2 3
1/2 3

=. f f (x )dx + j ~(3
25
1/2
- x )dx
1/.3
 1.36

c~ing l'dr:

-
9

Moments, Moment Generatin

g Function (mgf) and Prop
erties:
Expectation:

If Xi s a r.v., then the expectatio

n of X (or Mean), denoted by
E[X}, is defined by
,I

L XiP x (Xi) , if X is discrete

i
E (X1 =. oo

j xfx(x)dx, if Xis continuous

, - oo

E[X] is sometimes
possible values of Xdenhoted by X T .
. he expected , ere of~
' w ere each value is weightedvalue of X is a we1ghted, ,°' e~a~ value·
by the probability that )<.. takes
 Probability and RHn<lom Variables 1.37

and variance:
Moments . 1

ed by,
,, • •

, t of the r.v..\. denoted by /~[ Y ·], 1s defin

---~ ------- -----,
llt:ll - ✓
Thl fl th 111L1l ~
,L (.ri tPx (:r:;.), if X i5 disc rete
i
F[ Xll] == ◄ co

j x"fx(x)dx, i f Xis continuous

, -oo
th
The first moment, E[X ], is the expe cte~ value of X . The n central
f ir , =- 1. :?. 3....
l /
2
t IOf moment about the mean), denoted
by E[( X - xr ] and is defined by
1111_1111en ,

,L (xi - Xf Px (xi), if Xis discrete

i

E [(X - xr ] = ◄ 00

j (x - Xf f x(x)dx, if Xis continuous

-CX)

Note:
l. E [ (X - X)
2
] is called the variance of X and is denoted by oJ.
ed asµ~ = E [Xn ]
2. nth moment about the origin is, denoted byµ~, defin
ed as µn = E [ (X - X) n]
3. nth moment about the mean Xis , denoted µn, defin
4. µ~ = E[X ], the mean value of X
5. µ2 = E[(X - X)2], the variance of X. Also
2
µ2 = E[( X - X) ]
2
= E[X 2 - 2XX + X ]

= E[X 2] - 2X E[X ] + E[X 2]

2 2
= E[X 2] - 2X + X , since X = E [X]
2 2 2 2
= E[X ] - X = E[X ] - (E [X ])

6. Var(aX + b) = E[(aX + b)2] - (E[a X + b])2

== E[a2X2+ 2abX + b2] - [aE(X) + b]2

= a2E[X2] + 2abE(X) + b2 - a2[E(X)]2 - 2abE(.X) - b2

== a2 { E[X2) - [E(X)]2}
2
== a Var (X)
 1 .38 Proha/JilitJ and N 11 mh<'1' Thcoty

MGF:
. ·,,t1 ti.;, \ ' deno tc,.d by J\l ,x: (t), is <lefined as
11h' MGl·• ot, a r:mdl1 ill Yann

~ etxip(xi), if .X is a discrete r.v

I i
/ Mxtt) = E[(; t\]-
- 00

I
j etx J(x) dx, if X is a continuous r.v
-oo

Properties of MGF:

l. E[_Yn] = n! {Coefficientoftn in Nlx(t)}

Proof: We have

tX t X 2 2
lvlx(t) = E[etx] = E[l + 1! + 2! + ...]
t2 2
tn
= l + tE[X] + 21 E[X ] + ... + n !E[Xn] + ...
Therefore,

tn
E[Xn] = Coefficient of - in Nlx(t)
n.1
= n ! {Coefficientof tn in Mx(t)}

Proof: Differentiating n times both sides of equation (1 ) w.r.t 't'

dn
dtn J,.,fx(t) = E[Xn] + terms involving t and higher powers oft
""'° (::Mx(t)) ,~o = E[Xn]

3. IfY = aX +- b, then Mv(t) = ebtMx(at)

Proof:

Nly(t) = E[eYt] = E[et(aX+b)]

= E[ebt.eatX] = ebt E[eatX]
= e6tlvix(at)
 l'ml>:,l,ility ;,ruf U:indom Variables 1.39

\
1
),.. wlu;rc X :ind Varc 111dcpend cn1 vi11fablcs, lhcn tvf /,(t) = tvrx(t) fvfy (t)
it 1i' Z
proof:
,, (t) = 1£l<l'·] :: "✓ [,/CX l Y)l
= B[e'X .,~n, ] =- /~[, lX] f~[eLY],
sin ce X and Y ,ire independent
:.:} 1\!. (t) = 1'vf \'. (l) J\tfy (I)

problems:
2(1 - x) , 0<x<l
I. 1f the pdf of X is given by f (x) ={O , otherwise

2
(i) show that E[Xr] = (r + l)(r + 2)
evaluate E[(2X + 1)
2
(ii) ]

Solution:

0 (0) 1 (0)
j xr f(x)dx = j j xr J(x)dx + j
00 00

(i) E[xr] = xrfet'fclx + xrJW,x

-00 -00 0 1

-oo f(x)=O 0 1 f(x)=O

'---v--'
f(x)=2(1-x)

= j xr2(l - x)dx
0
1

=2 j (xr - xr+ 1
) dx
0

= 2 [ xr+ l
r+l
_ xr+2
r+2
l l

= 2[-1- __1_] - 2[(r + 2) - (r + 1)]

r+l r+2 - (r+l)(r+2)
2 (I)
:- -:------
(r -t- l)(r + 2)
 1

1 :1
2) 2( n
J \ (I thl

l dl. ~
2 :\ ( I) (,

\ \ 11 + l ){ i 4 !\)

l , O<x<k
1. Ar., X has the pdf given by f(x) === { k
0 , otherwise

Find (i) .MGE (ii) r th moment about the origin (iii) Mean and Variance.

Solution:

f(x)=O
f(x)=O

= E [etX]
00
oo O (0) k (0)

J
=_ etxf(x) dx = .I e "J0;'5dx + .I etx f(x)dx + j e "J¥idx
1
1

oo -oo O k

_J
-
k

o
l
etx kdx = -l ( _etx ) k = -(
k t
l it _ 1)
o kt
 ,--ronnnillly and Random Variabl!s

•·\ ,.th ttH' 111 cnl abnul the origin is

(II I

= l'l\'']
L: l () ( {) ) ~: X

- ./ /f(,r)d~· = / .r' f (.rfi.r I ./ / f (1),l,, 4 / .r'.J).<t'Jll

-(.~ -c.'O n f.

= l ,.
k

,l'
t l
k. ' .c
0

- l_ (.i:"+l )k= ~ h/+ l = ~ (

- k r+l O kr+l r+l

liii) Tak.ing ,. == 1 in (1),

k k
E[.X] =- =-
1+1 2
Taking r == 2 in (1),

Therefore,
k
Mean of X = E[X] =
2
and
2
Var(X) = E(X 2 ) - [E(X)]

= Gr=~2 - ~2 - : =~;
3. Let X be a r.v with E[X] =land E[X(X - l)] = 4. Find

Var ( ~) and Var(2 - 3X)

Solution:
Given that E[X] = 1, E[X(X - l)] =4
⇒ E(X 2 - X) = 4
⇒ E(X 2 ) - E(X) = 4 ⇒ E (X 2 )) - 1=4
⇒ E(X 2 ) =5
Therefore
'
Var(X) = E(X 2 ) - [E(X)] 2 =5- 12 =4
 ., \ lltlll1cr
l 1,c,H)
Ptnh:ilu I(,
\ (lJJU
•.\2

G) 2
v,,r(X) = ,J
~(il ) - I

. l) (4) = :w
., 3X)
. ( 3)2 wr(X)
.
om! Vnr(.., o l ot r =
']
, , 1£(X 1 - • O
\ nrc dehn~d b) ; .X > 2] - 0
4 If the monlt:,nts ot 8 P
t.\.\ 1] - u.G 1\IH.l ' ( -
. \ {1] 0. .

~olulion: 1 2 ;3. .. ·
V • o.G, for f == ., l
G\ en that E( Xr]

..'fg.f == 'h (t)

=- Ele' \ ]
I'()

= L etx P[X === x]

et P[X === 1] + e2t P[X = 2] + ...
= x=O
0
e P[X = 0] + 2t
= P[X = O] + p [ - .
X - 1] et+ P[X = 2]e + ... oo
OJ
Again.

Mx(t) = E[etX]
2 2
t X tX ]
-El+
[ - + - + ....
- 1! 2!
t t2
_
- E(l ) + -E(X
1! ) + -E(X
2!
2
) + ····
t t2
= 1 + 1! (0.6) + 2! (0.6) + ····
t t2
= 0.4 + 0.6 + 1! (0.6) + 2! (0.6) + ....
2
= 0.4 + 0.6 [1 + 1!t + t2! + .. ·]
= 0.4 + 0.6et
From (1) and (2),

P[X == O] + P[X == l]et + P [X = 2]e2t + ... = 0.4 + 0.6et

=;, P[X = 0] ..., 0.4, P[X = l] = 0.6, P[X = 2] = 0, P [X = 3] = O, ···
_,, P[X = OJ= 0.4, P[X = l] = 0.6, P[X = r] = 0, for r = 2.3, ...
 . probability Distributions
special
t
e Distributions
oiscre
Bernoulli trial :
. ·a1 is an experiment that results in one of the two outcomes: success and
.\ f3ernoull1 tn
failure.

_ Binomial Distribution:
1
. that we conduct n independent Bernoulli trials. Let p denote the probability of
Suppo~e
. •n a trial and let X denote the number of successes in n such trials. Then we say
succe~s 1 . . .
that X is a binomial variate and its pmf 1s given by

'.3"rx = x] = nCxpxqn- x , x = 0, 1, ... , n; where q = l - p. I

Toe real numbers n and p are called the parameters of the distribution.

Derivation of the MGF, Mean and Variance of a Binomial Distribution:

Solution:

_vfgf = lvl x(t) = E[etx]

n
= L etx P [X = x]
x=O
n
= L etxncxpxqn-x
x=O
n
= L ncx(pet?qn-x
x=O
= (q + petr
(Since: When n is a positive integer,

(a + br = an - nc1 an-lb - .... ~ bn

= ncoan-ObO + nc1an-lbl + ..... + ncnan-nbn
n
= L nexbx an-x)
x=O