Moment generating function (m.g.

f)
Moment generating function of a discrete random variable X about
a denoted by 𝑴𝒂 (𝒕) defined by
𝑀𝑎 (𝑡) = 𝐸[𝑒 𝑡(𝑥−𝑎) ]

= ∑ 𝑒 𝑡(𝑥𝑖 −𝑎) 𝑃𝑖

Expanding

𝑥
𝑥2 𝑥3
𝑒 = 1 + 𝑥 + + + ⋯ ..
2! 3!

2 3
(𝑡(𝑥𝑖 − 𝑎)) (𝑡(𝑥𝑖 − 𝑎))
𝑀𝑎 (𝑡) = ∑ 𝑃𝑖 [1 + 𝑡(𝑥𝑖 − 𝑎) + + + ⋯.]
2! 2!

𝑡2 𝑡3
= ∑ 𝑃𝑖 + ∑ 𝑡(𝑥𝑖 − 𝑎)𝑃𝑖 + ∑ (𝑥𝑖 − 𝑎) 𝑃𝑖 + ∑ (𝑥𝑖 − 𝑎)3 𝑃𝑖 + ⋯ …
2
2! 3!

𝑡2 𝑡3
= ∑ 𝑃𝑖 + 𝑡 ∑(𝑥𝑖 − 𝑎)𝑃𝑖 + ∑(𝑥𝑖 − 𝑎) 𝑃𝑖 + ∑(𝑥𝑖 − 𝑎)3 𝑃𝑖 + ⋯ …
2
2! 3!

′
𝑡2 ′ 𝑡3 ′
= 1 + 𝑡𝜇1 + 𝜇2 + 𝜇3 + ⋯ … …
2! 3!
𝑡𝑟
Thus, the coefficient of is the rth moment of X about a (𝜇𝑟 ′ ).in this way
𝑟!
𝑀𝑎 (𝑡) generates moments. Hence 𝑀𝑎 (𝑡) is called moment
generating function.

EM III_SMITA N 1
Moment generating function of a continuous random variable X
about a denoted by 𝑴𝒂 (𝒕) defined by

𝑀𝑎 (𝑡) = 𝐸[𝑒 𝑡(𝑥−𝑎) ]

𝑀𝑎 (𝑡) = ∫ 𝑒 𝑡(𝑥−𝑎) 𝑓(𝑥 )𝑑𝑥

−∞

∞ 2 3
(𝑡(𝑥𝑖 − 𝑎)) (𝑡(𝑥𝑖 − 𝑎))
= ∫ 𝑓(𝑥 ) [1 + 𝑡(𝑥𝑖 − 𝑎) + + + ⋯ . ] 𝑑𝑥
2! 2!
−∞

∞ ∞ ∞
𝑡2
= ∫ 𝑓(𝑥)𝑑𝑥 + ∫ 𝑡(𝑥𝑖 − 𝑎)𝑓(𝑥)𝑑𝑥 + ∫ (𝑥𝑖 − 𝑎)2 𝑓(𝑥)𝑑𝑥
2!
−∞ −∞ −∞
∞
3
𝑡
+ ∫ (𝑥𝑖 − 𝑎)3 𝑓(𝑥)𝑑𝑥 + ⋯
3!
−∞
∞ ∞
𝑡2
= 1 + 𝑡 ∫ (𝑥𝑖 − 𝑎)𝑓(𝑥)𝑑𝑥 + ∫ (𝑥𝑖 − 𝑎)2 𝑓(𝑥)𝑑𝑥
2!
−∞ −∞
∞
3
𝑡
+ ∫ (𝑥𝑖 − 𝑎)3 𝑓(𝑥)𝑑𝑥 + ⋯
3!
−∞

𝑡2 ′ 𝑡3 ′
′
= 1 + 𝑡𝜇1 + 𝜇2 + 𝜇3 + ⋯ … … (𝐴)
2! 3!

∞
∫−∞(𝑥𝑖 − 𝑎)𝑟 𝑓(𝑥)𝑑𝑥 is the rth moment of X about a (𝜇𝑟 ′ ).

EM III_SMITA N 2
To find moments of various orders from m.g.f

Differentiating equation (A) w.r.t t

′
𝑡2 ′
𝑡3
𝑀𝑎 (𝑡) = 1 + 𝑡𝜇1 + 𝜇2 + 𝜇3 ′ + ⋯ … …
2! 3!

𝑑 ′
2𝑡 ′ 3𝑡 2 ′
[𝑀𝑎 (𝑡)] = 𝜇1 + 𝜇2 + 𝜇 + ⋯……
𝑑𝑡 2! 3! 3

𝑑 ′ ′
𝑡2 ′
[𝑀𝑎 (𝑡)] = 𝜇1 + 𝑡𝜇2 + 𝜇3 + ⋯ … …
𝑑𝑡 2!

Putting t=0
𝑑
[𝑀𝑎 (𝑡)]𝑡=0 = 𝜇1 ′
𝑑𝑡

𝑑2 ′
2𝑡 ′
[ 𝑀𝑎 ( 𝑡)] = 𝜇2 + 𝜇 + ⋯……
𝑑𝑡 2 2! 3

𝑑2
2
[𝑀𝑎 (𝑡)]𝑡=0 = 𝜇2 ′
𝑑𝑡

𝑑3
3
[𝑀𝑎 (𝑡)]𝑡=0 = 𝜇3 ′
𝑑𝑡

𝑑𝑟
𝑟
[𝑀𝑎 (𝑡)]𝑡=0 = 𝜇𝑟 ′
𝑑𝑡

EM III_SMITA N 3
Moment generating function of a discrete random variable X about
origin denoted by 𝑴𝟎 (𝒕) defined by
𝑀0 (𝑡) = 𝐸[𝑒 𝑡(𝑥−0) ]

= ∑ 𝑒 𝑡𝑥𝑖 𝑃𝑖

Moment generating function of a discrete random variable X about

̅ denoted by 𝑴𝒂 (𝒕) defined by
the mean 𝒙
𝑀𝒙̅ (𝑡) = 𝐸[𝑒 𝑡(𝑥−𝒙̅) ]

= ∑ 𝑒 𝑡(𝑥𝑖 −𝒙̅) 𝑃𝑖

Expanding

𝑥
𝑥2 𝑥3
𝑒 = 1 + 𝑥 + + + ⋯ ..
2! 3!
2 3
(𝑡(𝑥𝑖 − 𝒙
̅)) (𝑡(𝑥𝑖 − 𝒙
̅))
𝑀𝒙̅ (𝑡) = ∑ 𝑃𝑖 [1 + 𝑡(𝑥𝑖 − 𝒙
̅) + + + ⋯.]
2! 2!

𝑡2 2
𝑡3
= ∑ 𝑃𝑖 + ∑ 𝑡(𝑥𝑖 − 𝒙
̅ )𝑃𝑖 + ∑ (𝑥𝑖 − 𝒙 ̅ )3 𝑃𝑖 + ⋯ …
̅ ) 𝑃𝑖 + ∑ (𝑥𝑖 − 𝒙
2! 3!

𝑡2 2
𝑡3
= ∑ 𝑃𝑖 + 𝑡 ∑(𝑥𝑖 − 𝒙
̅ )𝑃𝑖 + ∑(𝑥𝑖 − 𝒙 ̅ )3 𝑃𝑖 + ⋯ …
̅ ) 𝑃𝑖 + ∑(𝑥𝑖 − 𝒙
2! 3!

𝑡2 𝑡3
= 1 + 𝑡𝜇1 + 𝜇2 + 𝜇3 + ⋯ … …
2! 3!
𝑡𝑟
Thus, the coefficient of ̅ 𝜇𝑟 .
is the rth moment of X about𝒙
𝑟!

EM III_SMITA N 4
 1) A random variable X has the following probability distribution

𝑥 0 1 2
𝑃(𝑋 = 𝑥) 1/3 1/3 1/3

Find the moment generating function and hence find the mean and Variance.
Solution

𝑀0 (𝑡) = 𝐸[𝑒 𝑡𝑥 ]

= ∑ 𝑒 𝑡𝑥𝑖 𝑃𝑖

1 1 1
= 𝑒 𝑡(0) + 𝑒 𝑡(1) + 𝑒 𝑡(2)
3 3 3

1
= [1 + 𝑒 𝑡 + 𝑒 2𝑡 ]
3

𝑑 1 2𝑡
[𝑀0 (𝑡)] = [𝑒𝑡 + 2𝑒 ]
𝑑𝑡 3

𝑑 1
𝜇1 ′ = [𝑀0 (𝑡)]𝑡=0 = [1 + 2] = 1
𝑑𝑡 3

𝑑2 1 𝑡 2𝑡
[ 𝑀0 ( 𝑡)] = [ 𝑒 + 4𝑒 ]
𝑑𝑡 2 3

′
𝑑2 1 5
𝜇2 = 2 [𝑀0 (𝑡)]𝑡=0 = [1 + 4] =
𝑑𝑡 3 3

EM III_SMITA N 5
 𝜇1 ′ = 𝑚𝑒𝑎𝑛 = 1

2
𝑉𝑎𝑟𝑖𝑎𝑛𝑐𝑒 𝑖𝑠 𝜇2 = 𝜇2 ′ − 𝜇1 ′

5
= −1
3

2
=
3

2)A random variable X has the following probability distribution

𝑥 0 1 2 3
𝑃(𝑋 = 𝑥) 1/6 1/3 1/3 1/6

Find the moment generating function and hence find the first four raw
moments and first four central moments.
Solution

𝑀0 (𝑡) = 𝐸[𝑒 𝑡𝑥 ]

= ∑ 𝑒 𝑡𝑥𝑖 𝑃𝑖

1 1 1 1
= 𝑒 𝑡(0) + 𝑒 𝑡(1) + 𝑒 𝑡(2) + 𝑒 𝑡(3)
6 3 3 6

1 1 𝑡 1 2𝑡 1 3𝑡
= + 𝑒 + 𝑒 + 𝑒
6 3 3 6

EM III_SMITA N 6
First four Raw moments

𝑑 1 2 3
[𝑀0 (𝑡)] = 𝑒𝑡 + 𝑒2𝑡 + 𝑒3𝑡
𝑑𝑡 3 3 6

𝑑 1 2 1
[𝑀0 (𝑡)] = 𝑒𝑡 + 𝑒2𝑡 + 𝑒3𝑡
𝑑𝑡 3 3 2

𝑑 1 2 1 3
𝜇1 ′ = [𝑀0 (𝑡)]𝑡=0 = + + =
𝑑𝑡 3 3 2 2

𝑑2 1 𝑡 4 2𝑡 3 3𝑡
[ 𝑀0 ( 𝑡)] = 𝑒 + 𝑒 + 𝑒
𝑑𝑡 2 3 3 2

′
𝑑2 1 4 3 19
𝜇2 = 2 [𝑀0 (𝑡)]𝑡=0 = + + =
𝑑𝑡 3 3 2 6

𝑑3 1 𝑡 8 2𝑡 9 3𝑡
[ 𝑀0 ( 𝑡)] = 𝑒 + 𝑒 + 𝑒
𝑑𝑡 3 3 3 2

′
𝑑3 1 8 9 15
𝜇3 = 3 [𝑀0 (𝑡)]𝑡=0 = + + =
𝑑𝑡 3 3 2 2

𝑑4 1 𝑡 16 2𝑡 27 3𝑡
[ 𝑀0 ( 𝑡)] = 𝑒 + 𝑒 + 𝑒
𝑑𝑡 4 3 3 2

′
𝑑4 1 16 27 115
𝜇4 = 4 [𝑀0 (𝑡)]𝑡=0 = + + =
𝑑𝑡 3 3 2 6

EM III_SMITA N 7
First four Central moments
𝜇1 = 0

′2
19 3 2 11
𝜇2 = 𝜇2 ′ − 𝜇1 = −( ) = = 0.9166
6 2 12

3
𝜇3 = 𝜇3 ′ − 3𝜇2 ′ 𝜇1 ′ + 2𝜇1 ′
15 19 3 3 3
= −3× × + 2( ) = 0
2 6 2 2

′2 4
𝜇4 = 𝜇4 ′ − 4𝜇3 ′ 𝜇1 ′ + 6𝜇2 ′ 𝜇1 − 3𝜇1 ′
115 15 3 19 3 2 3 4
= −4× × +6× ×( ) −3×( )
6 2 2 6 2 2
83
= = 1.72913
48

𝟑) 𝐹𝑖𝑛𝑑 𝑡ℎ𝑒 𝑚. 𝑔. 𝑓 𝑜𝑓 𝑡ℎ𝑒 𝑓𝑜𝑙𝑙𝑜𝑤𝑖𝑛𝑔 𝑑𝑖𝑠𝑡𝑟𝑖𝑏𝑢𝑡𝑖𝑜𝑛

𝑥 -2 3 1
𝑃(𝑋 = 𝑥) 1/3 1/2 1/6

Hence find first four central moments

Solution

𝑀0 (𝑡) = 𝐸[𝑒 𝑡𝑥 ]
= ∑ 𝑒 𝑡𝑥𝑖 𝑃𝑖
1 1 1
= 𝑒 𝑡(−2) + 𝑒 𝑡(3) + 𝑒 𝑡(1)
3 2 6

EM III_SMITA N 8
 1 −2𝑡 1 3𝑡 1 𝑡
= 𝑒 + 𝑒 + 𝑒
3 2 6

First four Raw moments

𝑑 −2 −2𝑡 3 3𝑡 1 𝑡
[𝑀0 (𝑡)] = 𝑒 + 𝑒 + 𝑒
𝑑𝑡 3 2 6

𝑑 −2 3 1
𝜇1 ′ = [𝑀0 (𝑡)]𝑡=0 = + + =1
𝑑𝑡 3 2 6

𝑑2 4 −2𝑡 9 3𝑡 1 𝑡
[ 𝑀0 ( 𝑡)] = 𝑒 + 𝑒 + 𝑒
𝑑𝑡 2 3 2 6

′
𝑑2 4 9 1
𝜇2 = 2 [𝑀0 (𝑡)]𝑡=0 = + + = 6
𝑑𝑡 3 2 6

𝑑3 −8 −2𝑡 27 3𝑡 1 𝑡
[ 𝑀0 ( 𝑡)] = 𝑒 + 𝑒 + 𝑒
𝑑𝑡 3 3 2 6

′
𝑑3 −8 27 1
𝜇3 = 3 [𝑀0 (𝑡)]𝑡=0 = + + = 11
𝑑𝑡 3 2 6

𝑑4 16 −2𝑡 81 3𝑡 1 𝑡
[ 𝑀0 ( 𝑡)] = 𝑒 + 𝑒 + 𝑒
𝑑𝑡 4 3 2 6

′
𝑑4 16 81 1
𝜇4 = 4 [𝑀0 (𝑡)]𝑡=0 = + + = 46
𝑑𝑡 3 2 6

EM III_SMITA N 9
First four Central moments
𝜇1 = 0

2
𝜇2 = 𝜇2 ′ − 𝜇1 ′ = 6 − (1)2 = 5

3
𝜇3 = 𝜇3 ′ − 3𝜇2 ′ 𝜇1 ′ + 2𝜇1 ′

= 11 − 3 × 6 × 1 + 2(1)3 = −5

′2 4
𝜇4 = 𝜇4 ′ − 4𝜇3 ′ 𝜇1 ′ + 6𝜇2 ′ 𝜇1 − 3𝜇1 ′

= 46 − 4 × 11 × 1 + 6 × 6 × (1)2 − 3 × (1)4 = 35

Second Method

𝑀𝒙̅ (𝑡) = 𝐸[𝑒 𝑡(𝑥−𝒙̅) ]

= ∑ 𝑒 𝑡(𝑥𝑖 −𝒙̅) 𝑃𝑖

̅ = ∑ 𝒙𝒊 𝑷𝒊
𝒙

1 1 1
= −2 × + 3 × + 1 × = 1
3 2 6

EM III_SMITA N 10
 𝑀𝒙̅ (𝑡) = ∑ 𝑒 𝑡(𝑥𝑖 −𝒙̅) 𝑃𝑖

1 1 1
= 𝑒 𝑡(−2−1) + 𝑒 𝑡(3−1) + 𝑒 𝑡(1−1)
3 2 6

1 1 1
= 𝑒 −3𝑡 + 𝑒 2𝑡 +
3 2 6

𝑑 −3 −3𝑡 2 2𝑡
[𝑀𝑥̅ (𝑡)] = 𝑒 + 𝑒
𝑑𝑡 3 2

𝑑
𝜇1 = [𝑀𝑥̅ (𝑡)]𝑡=0 = −1 + 1 = 0
𝑑𝑡

𝑑2
[𝑀𝒙̅ (𝑡)] = 3𝑒−3𝑡 + 2𝑒2𝑡
𝑑𝑡 2

𝑑2
𝜇2 = 2 [𝑀𝒙̅ (𝑡)]𝑡=0 = 3 + 2 = 5
𝑑𝑡

𝑑3
[𝑀𝒙̅ (𝑡)] = −9𝑒−3𝑡 + 4𝑒2𝑡
𝑑𝑡 3

𝑑3
𝜇3 = 3 [𝑀𝒙̅ (𝑡)]𝑡=0 = −9 + 4 = −5
𝑑𝑡

𝑑4
[𝑀𝒙̅ (𝑡)] = 27𝑒−3𝑡 + 8𝑒2𝑡
𝑑𝑡 4

EM III_SMITA N 11
 𝑑4
𝜇4 = 4 [𝑀𝒙̅ (𝑡)]𝑡=0 = 27 + 8 = 35
𝑑𝑡

1
4) A Random variable X has probability density function , 𝑥 = 1,2,3 ….
2𝑥
𝐹𝑖𝑛𝑑 𝑚. 𝑔. 𝑓 𝑎𝑛𝑑 ℎ𝑒𝑛𝑐𝑒 𝑓𝑖𝑛𝑑 𝑚𝑒𝑎𝑛 𝑎𝑛𝑑 𝑣𝑎𝑟𝑖𝑎𝑛𝑐𝑒.

Solution
1
𝑃(𝑥 = 𝑥) = , 𝑥 = 1,2,3 ….
2𝑥
𝑀0 (𝑡) = 𝐸[𝑒 𝑡𝑥 ]
= ∑ 𝑒 𝑡𝑥𝑖 𝑃𝑖
1
= ∑ 𝑒 𝑡𝑥𝑖
2𝑥

1 1 1
= 𝑒 𝑡(1) + 𝑒 𝑡(2) 2 + 𝑒 𝑡(3) 3 + ⋯ ..
2 2 2

𝑒 𝑡 𝑒 2𝑡 𝑒 3𝑡
= + 2 + 3 + ⋯ ..
2 2 2

2
𝑒𝑡 𝑒𝑡 𝑒𝑡
= [1 + + ( ) + ⋯ . ]
2 2 2

(1 − 𝑥)−1 = 1 + 𝑥 + 𝑥 2 + 𝑥 3 + ⋯ ..

−1
𝑒𝑡 𝑒𝑡
= (1 − )
2 2

EM III_SMITA N 12
 𝑒𝑡 1
=
2 𝑒𝑡
(1 − )
2
𝑒𝑡 2
=
2 2 − 𝑒𝑡

𝑒𝑡
=
2 − 𝑒𝑡

𝑑 (2 − 𝑒𝑡 )𝑒𝑡 − 𝑒𝑡 (−𝑒𝑡 )
[𝑀0 (𝑡)] = 2
𝑑𝑡 (2 − 𝑒𝑡 )

2𝑒 𝑡 − 𝑒 2𝑡 + 𝑒 2𝑡
=
(2 − 𝑒 𝑡 )2

2𝑒 𝑡
=
(2 − 𝑒 𝑡 )2

𝑑 2
𝜇1 ′ = [𝑀0 (𝑡)]𝑡=0 = =2
𝑑𝑡 ( 2 − 1) 2

𝑡
𝑑2 (2 − 𝑒𝑡 )2 2𝑒𝑡 − 2𝑒 2(2−𝑒𝑡 )(−𝑒𝑡 )
[𝑀0 (𝑡)] = 4
𝑑𝑡 2 (2 − 𝑒𝑡 )
(2−𝑒 𝑡 )2𝑒 𝑡 − 2𝑒 𝑡 2(−𝑒 𝑡 )
=
(2 − 𝑒 𝑡 )3

EM III_SMITA N 13
 ′
𝑑2 ( 2 − 1) 2 + 4
𝜇2 = 2 [𝑀0 (𝑡)]𝑡=0 = 3
=6
𝑑𝑡 (2 − 1)

𝜇1 ′ = 𝑚𝑒𝑎𝑛 = 2

2
𝑉𝑎𝑟𝑖𝑎𝑛𝑐𝑒 𝑖𝑠 𝜇2 = 𝜇2 ′ − 𝜇1 ′

=6−4

=2

5) If X denotes the outcome when a fair die is tossed, Find the m.g.f
of X and hence find the mean and variance of X
𝑥 1 2 3 4 5 6
𝑃(𝑋 = 𝑥) 1/6 1/6 1/6 1/6 1/6 1/6

Solution

𝑀0 (𝑡) = 𝐸[𝑒 𝑡𝑥 ]

= ∑ 𝑒 𝑡𝑥𝑖 𝑃𝑖

1 1 1 1 1 1
= 𝑒 𝑡(1) + 𝑒 𝑡(2) + 𝑒 𝑡(3) + 𝑒 𝑡(4) + 𝑒 𝑡(5) + 𝑒 𝑡(6)
6 6 6 6 6 6

1
= [𝑒 𝑡 + 𝑒 2𝑡 + 𝑒 3𝑡 + 𝑒 4𝑡 + 𝑒 5𝑡 + 𝑒 6𝑡 ]
6

EM III_SMITA N 14
 𝑑 1 3𝑡 4𝑡 6𝑡
[𝑀0 (𝑡)] = [𝑒𝑡 + 2𝑒2𝑡 + 3𝑒 + 4𝑒 + 5𝑒5𝑡 + 6𝑒 ]
𝑑𝑡 6

𝑑 1 21 7
𝜇1 ′ = [𝑀0 (𝑡)]𝑡=0 = [1 + 2 + 3 + 4 + 5 + 6] = =
𝑑𝑡 6 6 2

𝑑2 1 𝑡 3𝑡 4𝑡 6𝑡
[ 𝑀0 ( 𝑡)] = [𝑒 + 4𝑒2𝑡 + 9𝑒 + 16𝑒 + 25𝑒5𝑡 + 36𝑒 ]
𝑑𝑡 2 6

′
𝑑2 1 91
𝜇2 = 2 [𝑀0 (𝑡)]𝑡=0 = [1 + 4 + 9 + 16 + 25 + 36] =
𝑑𝑡 6 6

7
𝜇1 ′ = 𝑚𝑒𝑎𝑛 =
2

2
𝑉𝑎𝑟𝑖𝑎𝑛𝑐𝑒 𝑖𝑠 𝜇2 = 𝜇2 ′ − 𝜇1 ′

91 7 2
= −( )
6 2

35
=
12

6) Find the m.g.f of a random variable X if the 𝑟𝑡ℎ moment about the
origin is 𝑟!
Solution

𝝁𝒓 ′ = 𝑬[𝒙𝒓 ] = 𝒓!
EM III_SMITA N 15
 𝑬(𝒙) = 𝟏!

𝑬(𝒙𝟐 ) = 𝟐!

𝑬(𝒙𝟑 ) = 𝟑!

𝑀0 (𝑡) = 𝐸[𝑒 𝑡𝑥 ]

𝑥
𝑥2 𝑥3
𝑒 = 1 + 𝑥 + + + ⋯ ..
2! 3!

𝑡2𝑥2 𝑡3𝑥3
𝑀0 (𝑡) = 𝐸[1 + 𝑡𝑥 + + + ⋯..]
2! 3!

𝑡2𝑥2 𝑡3𝑥3
= 𝐸[1] + 𝐸[𝑡𝑥] + 𝐸 [ ]+𝐸[ ] + ⋯ ..
2! 3!

𝑡2 𝑡3
= 1 + 𝑡𝐸[𝑥] + 𝐸 [𝑥 ] + + 𝐸 [𝑥 3 ] + ⋯ ..
2
2! 3!

𝑡2 𝑡3
= 1 + 𝑡 + 2! + + 3! + ⋯ ..
2! 3!

= 1 + 𝑡 + 𝑡 2 + 𝑡 3 + ⋯ ..

(1 − 𝑥)−1 = 1 + 𝑥 + 𝑥 2 + 𝑥 3 + ⋯ ..

1
𝑀0 (𝑡) = (1 − 𝑡)−1 =
1−𝑡

EM III_SMITA N 16
 1
𝑀0 (𝑡) =
1−𝑡

7) A random variable X has the probability distribution

1
𝑃(𝑋 = 𝑥 ) = 3𝐶𝑥 , 𝑥 = 0,1,2,3
8
Find the moment generating function of X and hence find mean and
variance.
Solution
1
𝑃(𝑋 = 𝑥 ) = 3𝐶𝑥
8

1 1
𝑃(𝑋 = 0) = 3𝐶0 =
8 8

1 3
𝑃(𝑋 = 1) = 3𝐶1 =
8 8

1 3
𝑃(𝑋 = 2) = 3𝐶2 =
8 8

1 1
𝑃(𝑋 = 3) = 3𝐶3 =
8 8

𝑥 0 1 2 3
𝑃(𝑋 = 𝑥) 1/8 3/8 3/8 1/8

𝑀0 (𝑡) = 𝐸[𝑒 𝑡𝑥 ]

= ∑ 𝑒 𝑡𝑥𝑖 𝑃𝑖

1 3 3 1
= 𝑒 𝑡(0) + 𝑒 𝑡(1) + 𝑒 𝑡(2) + 𝑒 𝑡(3)
8 8 8 8

EM III_SMITA N 17
 1
= [1 + 3𝑒 𝑡 + 3𝑒 2𝑡 + 𝑒 3𝑡 ]
8

𝑑 1 𝑡 3𝑡
[𝑀0 (𝑡)] = [3𝑒 + 6𝑒2𝑡 + 3𝑒 ]
𝑑𝑡 8

𝑑 1 12 3
𝜇1 ′ = [𝑀0 (𝑡)]𝑡=0 = [3 + 6 + 3] = =
𝑑𝑡 8 8 2

𝑑2 1 𝑡 3𝑡
[ 𝑀0 ( 𝑡)] = [3𝑒 + 12𝑒2𝑡 + 9𝑒 ]
𝑑𝑡 2 8

′
𝑑2 1
𝜇2 = 2 [𝑀0 (𝑡)]𝑡=0 = [3 + 12 + 9] = 3
𝑑𝑡 8

3
𝜇1 ′ = 𝑚𝑒𝑎𝑛 =
2

2
𝑉𝑎𝑟𝑖𝑎𝑛𝑐𝑒 𝑖𝑠 𝜇2 = 𝜇2 ′ − 𝜇1 ′

𝟑 𝟐
=3−( )
𝟐

3
=
4

EM III_SMITA N 18
 2
8) A random variable X has m.g.f given by 𝑀0 (𝑡) = .Find the
2−𝑡
standard deviation of X.

Solution

𝑑 (2 − 𝑡)(0) − 2(−1) 2
[𝑀0 (𝑡)] = 2
= 2
𝑑𝑡 (2 − 𝑡) (2 − 𝑡)

𝑑 2 2 1
𝜇1 ′ = [𝑀0 (𝑡)]𝑡=0 = = =
𝑑𝑡 (2 − 0)
2 4 2

𝑑2 (2 − 𝑡)2 (0) − 2(2)(2 − 𝑡)(−1)

[𝑀0 (𝑡)] = 4
𝑑𝑡 2 (2 − 𝑡)
4(2 − 𝑡) 4
= =
(2 − 𝑡)4 (2 − 𝑡)3

′
𝑑2 4 4 1
𝜇2 = 2 [𝑀0 (𝑡)]𝑡=0 = = =
𝑑𝑡 (2 − 0)
3 8 2

2
𝑉𝑎𝑟𝑖𝑎𝑛𝑐𝑒 𝑖𝑠 𝜇2 = 𝜇2 ′ − 𝜇1 ′

1 1 2
= −( )
2 2

1
=
4

EM III_SMITA N 19
 1
𝑺𝒕𝒂𝒏𝒅𝒂𝒓𝒅 𝒅𝒆𝒗𝒊𝒂𝒕𝒊𝒐𝒏 =
2

3
9) A random variable X has m.g.f given by 𝑀0 (𝑡) = .Obtain
3−𝑡
the mean and the standard deviation of X.

Solution
𝑑 (3 − 𝑡)(0) − 3(−1) 3
[𝑀0 (𝑡)] = 2
= 2
𝑑𝑡 (3 − 𝑡) (3 − 𝑡)

𝑑 3 3 1
𝜇1 ′ = [𝑀0 (𝑡)]𝑡=0 = = =
𝑑𝑡 (3 − 0)
2 9 3

𝑑2 (3 − 𝑡)2 (0) − 3(2)(3 − 𝑡)(−1)

[𝑀0 (𝑡)] = 4
𝑑𝑡 2 (3 − 𝑡)

6(3 − 𝑡) 6
= =
(3 − 𝑡)4 (3 − 𝑡)3

′
𝑑2 6 6 2
𝜇2 = 2 [𝑀0 (𝑡)]𝑡=0 = = =
𝑑𝑡 (3 − 0)
3 27 9

2
𝑉𝑎𝑟𝑖𝑎𝑛𝑐𝑒 𝑖𝑠 𝜇2 = 𝜇2 ′ − 𝜇1 ′

2 1 2
= −( )
9 3

EM III_SMITA N 20
 1
=
9
1
𝑆𝑡𝑎𝑛𝑑𝑎𝑟𝑑 𝑑𝑒𝑣𝑖𝑎𝑡𝑖𝑜𝑛 =
3

10) A random variable takes values X=0,1 with probabilities 𝑞 𝑎𝑛𝑑 𝑝

respectively such that 𝑞 + 𝑝 = 1.Find the moment generating
function of X and show that all moments about the origin are equal
to p.
Solution

𝑥 0 1
𝑃(𝑋 = 𝑥) 𝑞 𝑝

𝑀0 (𝑡) = 𝐸[𝑒 𝑡𝑥 ]

= ∑ 𝑒 𝑡𝑥𝑖 𝑃𝑖

= 𝑒 𝑡(0) 𝑞 + 𝑒 𝑡(1) 𝑝

𝑞+ 𝑝=1

𝑞 =1−𝑝

𝑀0 (𝑡) = 𝑒 𝑡(0) (1 − 𝑝) + 𝑒 𝑡(1) 𝑝

EM III_SMITA N 21
 = (1 − 𝑝) + 𝑒 𝑡(1) 𝑝

= 1 + (𝑒 𝑡 − 1)𝑝

𝑑
[𝑀0 (𝑡)] = 𝑒𝑡 𝑝
𝑑𝑡

𝑑
𝜇1 ′ = [𝑀0 (𝑡)]𝑡=0 = 𝑝
𝑑𝑡

𝑑2
2
[𝑀0 (𝑡)] = 𝑒𝑡 𝑝
𝑑𝑡

′
𝑑2
𝜇2 = 2 [𝑀0 (𝑡)]𝑡=0 = 𝑝
𝑑𝑡

𝑑3
3
[𝑀0 (𝑡)] = 𝑒𝑡 𝑝
𝑑𝑡

′
𝑑3
𝜇3 = 3 [𝑀0 (𝑡)]𝑡=0 = 𝑝
𝑑𝑡

11) A continuous random variable X has the probability distribution

4
𝑓(𝑥 ) = 𝑥 (9 − 𝑥 2 )𝑤ℎ𝑒𝑟𝑒 0 ≤ 𝑥 ≤ 3 𝑎𝑛𝑑 𝑓 (𝑥 ) = 0 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒
81
Find the moment generating function and hence the first four
moments about the origin.

EM III_SMITA N 22
Solution

𝑀0 (𝑡) = 𝐸[𝑒 𝑡𝑥 ]

3
4
= ∫ 𝑒 𝑡𝑥 𝑥 (9 − 𝑥 2 )𝑑𝑥
81
0

3
4
= ∫ 𝑒 𝑡𝑥 (9𝑥 − 𝑥 3 )𝑑𝑥
81
0

4 3
𝑒 𝑡𝑥 2
𝑒 𝑡𝑥 𝑒 𝑡𝑥
= [(9𝑥 − 𝑥 ) ( ) − (9 − 3𝑥 ) ( 2 ) + (−6𝑥 ) ( 3 )
81 𝑡 𝑡 𝑡
3
𝑒 𝑡𝑥
− (−6) ( 4 )]
𝑡 0

4 𝑒 3𝑡 𝑒 3𝑡 𝑒 3𝑡
= [−(9 − 27) ( 2 ) + (−18) ( 3 ) − (−6) ( 4 )
81 𝑡 𝑡 𝑡
1 1
+ (9) ( 2 ) − (6) ( 4 )]
𝑡 𝑡

4 𝑒 3𝑡 𝑒 3𝑡 𝑒 3𝑡 1 1
= [18 ( 2 ) − 18 ( 3 ) + 6 ( 4 ) + (9) ( 2 ) − (6) ( 4 )]
81 𝑡 𝑡 𝑡 𝑡 𝑡

𝝁𝒓 ′ = 𝑬[𝑿𝒓 ]

𝝁𝟏 ′ = 𝑬[𝑿]

EM III_SMITA N 23
 3
4
= ∫𝑥 𝑥 (9 − 𝑥 2 )𝑑𝑥
81
0

3
4
= ∫(9𝑥 2 − 𝑥 4 )𝑑𝑥
81
0

3
4 𝑥3 𝑥5
= [9 ( ) − ( )]
81 3 5 0

4 162 8
= × =
81 5 5

𝝁𝟐 ′ = 𝑬[𝑿𝟐 ]

3
4
= ∫ 𝑥2 𝑥 (9 − 𝑥 2 )𝑑𝑥
81
0

3
4
= ∫(9𝑥 3 − 𝑥 5 )𝑑𝑥
81
0

3
4 𝑥4 𝑥6
= [9 ( ) − ( )]
81 4 6 0

4 243
= × =3
81 4

EM III_SMITA N 24
 𝝁𝟑 ′ = 𝑬[𝑿𝟑 ]

3
4
= ∫ 𝑥3 𝑥 (9 − 𝑥 2 )𝑑𝑥
81
0

3
4
= ∫(9𝑥 4 − 𝑥 6 )𝑑𝑥
81
0

3
4 𝑥5 𝑥7
= [9 ( ) − ( )]
81 5 7 0

4 243 216
= × =
81 4 35

𝝁𝟒 ′ = 𝑬[𝑿𝟒 ]

3
4
= ∫ 𝑥4 𝑥 (9 − 𝑥 2 )𝑑𝑥
81
0

3
4
= ∫(9𝑥 5 − 𝑥 7 )𝑑𝑥
81
0

3
4 𝑥6 𝑥8 27
= [9 ( ) − ( )] =
81 6 8 0 2

EM III_SMITA N 25
 12) A random variable X has the following probability density
1 0<𝑥<1
function 𝑓(𝑥 ) = {
0 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒
Find the 𝑚. 𝑔. 𝑓. 𝜇𝑟 ′ , 𝑚𝑒𝑎𝑛 𝑎𝑛𝑑 𝑣𝑎𝑟𝑖𝑎𝑛𝑐𝑒

Solution

𝑀0 (𝑡) = 𝐸[𝑒 𝑡𝑥 ]

= ∫ 𝑒 𝑡𝑥 𝑓(𝑥 )𝑑𝑥
0

= ∫ 𝑒 𝑡𝑥 (1)𝑑𝑥
0

1
𝑒 𝑡𝑥
=[ ]
𝑡 0

1 𝑡
= [ 𝑒 − 1]
𝑡

𝑥
𝑥2 𝑥3
𝑒 = 1 + 𝑥 + + + ⋯ ..
2! 3!

1 𝑡2 𝑡3
= [1 + 𝑡 + + + ⋯ − 1]
𝑡 2! 3!

EM III_SMITA N 26
 1 𝑡2 𝑡3
= [𝑡 + + + ⋯ ]
𝑡 2! 3!

𝑡 𝑡2
1+ + +⋯
2! 3!

′
𝑡𝑟 1
𝜇𝑟 = 𝑐𝑜𝑒𝑓𝑓𝑖𝑐𝑖𝑒𝑛𝑡 𝑜𝑓 =
𝑟! (𝑟 + 1)

1
𝑚𝑒𝑎𝑛 = 𝜇1 ′ =
(1 + 1)
1
=
2
1
𝜇2 ′ =
(2 + 1)
1
=
3

2
𝑉𝑎𝑟𝑖𝑎𝑛𝑐𝑒 𝑖𝑠 𝜇2 = 𝜇2 ′ − 𝜇1 ′

1 1 2
= −( )
3 2

1
=
12

EM III_SMITA N 27
 13) A random variable X has the following probability density
−𝑘𝑥
( )
function 𝑓 𝑥 = { 𝑘𝑒 𝑥>0 𝑘>0
0 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒
Find the 𝑚. 𝑔. 𝑓, 𝑚𝑒𝑎𝑛 𝑎𝑛𝑑 𝑣𝑎𝑟𝑖𝑎𝑛𝑐𝑒

Solution

𝑀0 (𝑡) = 𝐸[𝑒 𝑡𝑥 ]
∞

= ∫ 𝑒 𝑡𝑥 𝑓(𝑥 )𝑑𝑥
0

= ∫ 𝑒 𝑡𝑥 𝑘𝑒 −𝑘𝑥 𝑑𝑥
0

∞
𝑒 −(𝑡−𝑘)𝑥
= 𝑘[ ]
−(𝑡 − 𝑘) 0

𝑘
=
𝑘−𝑡

𝑑 (𝑘 − 𝑡)(0) − 𝑘(−1) 𝑘
[𝑀0 (𝑡)] = 2
= 2
𝑑𝑡 (𝑘 − 𝑡) (𝑘 − 𝑡)

𝑑 𝑘 1
𝜇1 ′ = [𝑀0 (𝑡)]𝑡=0 = 2
=
𝑑𝑡 (𝑘 − 0) 𝑘

EM III_SMITA N 28
 𝑑2 (𝑘 − 𝑡)2 (0) − 𝑘(2)(𝑘 − 𝑡)(−1)
[𝑀0 (𝑡)] = 4
𝑑𝑡 2 (𝑘 − 𝑡)

2𝑘(𝑘 − 𝑡) 2𝑘
= =
(𝑘 − 𝑡)4 (𝑘 − 𝑡)3

′
𝑑2 2𝑘 2𝑘 2
𝜇2 = 2 [𝑀0 (𝑡)]𝑡=0 = 3
= 3
= 2
𝑑𝑡 (𝑘 − 0) 𝑘 𝑘

1
𝜇1 ′ = 𝑚𝑒𝑎𝑛 =
𝑘

2
𝑉𝑎𝑟𝑖𝑎𝑛𝑐𝑒 𝑖𝑠 𝜇2 = 𝜇2 ′ − 𝜇1 ′

2 1 2
= 2−( )
𝑘 𝑘

2
=
𝑘2

14) Find the m.g.f of the random variable having probability density
function
𝑥 0≤𝑥≤1
𝑓(𝑥) = {2 − 𝑥 1 ≤ 𝑥 ≤ 2
0 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒

Solution

EM III_SMITA N 29
𝑀0 (𝑡) = 𝐸[𝑒 𝑡𝑥 ]

= ∫ 𝑒 𝑡𝑥 𝑓(𝑥 )𝑑𝑥
0

1 2

= ∫ 𝑒 𝑡𝑥 𝑥𝑑𝑥 + ∫ 𝑒 𝑡𝑥 (2 − 𝑥)𝑑𝑥
0 1

1 2
𝑒 𝑡𝑥 𝑒 𝑡𝑥 𝑒 𝑡𝑥 𝑒 𝑡𝑥
= [(𝑥 ) ( ) − (1) ( 2 )] + [(2 − 𝑥 ) ( ) − (−1) ( 2 )]
𝑡 𝑡 0
𝑡 𝑡 1

𝑒𝑡 𝑒𝑡 1 𝑒 2𝑡 𝑒𝑡 𝑒𝑡
= − 2 + 2 + + 2 − (2 − 1) ( ) − ( 2 )
𝑡 𝑡 𝑡 𝑡 𝑡 𝑡

𝑒 𝑡 𝑒 𝑡 1 𝑒 2𝑡 𝑒 𝑡 𝑒 𝑡
= − 2+ 2+ 2 − − 2
𝑡 𝑡 𝑡 𝑡 𝑡 𝑡

𝑒 2𝑡 1 2𝑒 𝑡
= 2 + 2− 2
𝑡 𝑡 𝑡

(𝑒 𝑡 − 1)2
=
𝑡2
2
𝑒𝑡 − 1
=( )
𝑡

EM III_SMITA N 30
 15) A random variable X has the following probability density
1
−1<𝑥 <2
function 𝑓(𝑥 ) = {3
0 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒
Find the 𝑚. 𝑔. 𝑓 𝑜𝑓 𝑋.
Solution

𝑀0 (𝑡) = 𝐸[𝑒 𝑡𝑥 ]

= ∫ 𝑒 𝑡𝑥 𝑓(𝑥 )𝑑𝑥
0

2
1
= ∫ 𝑒 𝑡𝑥 𝑑𝑥
3
−1

2
1 𝑒 𝑡𝑥
= [ ]
3 𝑡 −1

1 𝑒 2𝑡 𝑒 −𝑡
= [ − ]
3 𝑡 𝑡

𝑒 2𝑡 − 𝑒 −𝑡
=
3𝑡

EM III_SMITA N 31
 16) A random variable X has the following probability density
−(𝑥−5)
( )
function 𝑓 𝑥 = { 𝑒 𝑥≥5
0 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒
Find the 𝑚. 𝑔. 𝑓, 𝑚𝑒𝑎𝑛 𝑎𝑛𝑑 𝑣𝑎𝑟𝑖𝑎𝑛𝑐𝑒
Solution

𝑀0 (𝑡) = 𝐸[𝑒 𝑡𝑋 ]

= ∫ 𝑒 𝑡𝑥 𝑒 −(𝑥−5) 𝑑𝑥
5

= ∫ 𝑒 𝑡𝑥 𝑒 −𝑥 𝑒 5 𝑑𝑥
5

∞
5
𝑒 −(1−𝑡)𝑥
=𝑒 [ ]
−(1 − 𝑡) 5

𝑒 −(1−𝑡)5
5
=𝑒 [ ]
1−𝑡

𝑒 −5 𝑒 5𝑡
5
=𝑒 [ ]
1−𝑡

𝑒 5𝑡
=
1−𝑡

EM III_SMITA N 32
 5𝑡
𝑑 (1 − 𝑡) (5𝑒 ) − 𝑒5𝑡 (−1) 5𝑒5𝑡 − 5𝑡𝑒5𝑡 + 𝑒5𝑡
[𝑀0 (𝑡)] = 2
= 2
𝑑𝑡 (1 − 𝑡) (1 − 𝑡)
6𝑒5𝑡 − 5𝑡𝑒5𝑡
= 2
(1 − 𝑡)

𝑑
𝜇1 ′ = [𝑀0 (𝑡)]𝑡=0 = 6
𝑑𝑡

𝑑2
[𝑀0 (𝑡)]
𝑑𝑡 2
(1 − 𝑡)2 (30𝑒5𝑡 − 5(𝑒5𝑡 + 5𝑡𝑒5𝑡 ) − (6𝑒5𝑡 − 5𝑡𝑒5𝑡 )(2)(1 − 𝑡)(−1)
= 4
(1 − 𝑡)

′
𝑑2
𝜇2 = 2 [𝑀0 (𝑡)]𝑡=0 = 30 − 5 + 12 = 37
𝑑𝑡

𝜇1 ′ = 𝑚𝑒𝑎𝑛 = 6

2
𝑉𝑎𝑟𝑖𝑎𝑛𝑐𝑒 𝑖𝑠 𝜇2 = 𝜇2 ′ − 𝜇1 ′

= 37 − (6)2

=1

EM III_SMITA N 33
 17) Find the m.g.f of the random variable having the probability density
function. Also find the mean and variance.
1
𝑓(𝑥) = { 2 , − 1 ≤ 𝑥 < 1
0 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒

Solution

𝑀0 (𝑡) = 𝐸[𝑒 𝑡𝑥 ]

= ∫ 𝑒 𝑡𝑥 𝑓(𝑥 )𝑑𝑥
0

1
1
= ∫ 𝑒 𝑡𝑥 𝑑𝑥
2
−1

1
1 𝑒 𝑡𝑥
= [ ]
2 𝑡 −1

1 𝑒 𝑡 𝑒 −𝑡
= [ − ]
2 𝑡 𝑡

𝑒 𝑡 − 𝑒 −𝑡
=
2𝑡

𝑥
𝑥2 𝑥3
𝑒 = 1 + 𝑥 + + + ⋯ ..
2! 3!

EM III_SMITA N 34
 1 𝑡2 𝑡3 𝑡2 𝑡3
= [1 + 𝑡 + + + ⋯ − (1 − 𝑡 + − + ⋯ )]
2𝑡 2! 3! 2! 3!

1 𝑡2 𝑡3 𝑡2 𝑡3
= [1 + 𝑡 + + + ⋯ − 1 + 𝑡 − + + ⋯ ]
2𝑡 2! 3! 2! 3!

1 2𝑡 3 2𝑡 5
= [2𝑡 + + …]
2𝑡 3! 5!

𝑡2 𝑡4
1+ + +⋯
3! 5!

′
𝑡𝑟 1
𝜇𝑟 = 𝑐𝑜𝑒𝑓𝑓𝑖𝑐𝑖𝑒𝑛𝑡 𝑜𝑓 =
𝑟! (𝑟 + 1)

𝑚𝑒𝑎𝑛 = 𝜇1 ′ = 0

1
𝜇2 ′ =
3

2
𝑉𝑎𝑟𝑖𝑎𝑛𝑐𝑒 𝑖𝑠 𝜇2 = 𝜇2 ′ − 𝜇1 ′

1
= − 02
3

1
=
3

EM III_SMITA N 35
 18) Arandom variable X has the following probability density
−2𝑥
function 𝑓(𝑥 ) = {2𝑒 𝑥>0
0 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒
Find the 𝑚. 𝑔. 𝑓, 𝑚𝑒𝑎𝑛 𝑎𝑛𝑑 𝑣𝑎𝑟𝑖𝑎𝑛𝑐𝑒
Solution

𝑀0 (𝑡) = 𝐸[𝑒 𝑡𝑥 ]

= ∫ 𝑒 𝑡𝑥 𝑓(𝑥 )𝑑𝑥
0

= ∫ 𝑒 𝑡𝑥 2𝑒 −2𝑥 𝑑𝑥
0

∞
𝑒 −(𝑡−2)𝑥
= 2[ ]
−(𝑡 − 2) 0

2
=
2−𝑡

𝑑 (2 − 𝑡)(0) − 2(−1) 2
[𝑀0 (𝑡)] = 2
= 2
𝑑𝑡 (2 − 𝑡) (2 − 𝑡)

EM III_SMITA N 36
 𝑑 2 1
𝜇1 ′ = [𝑀0 (𝑡)]𝑡=0 = =
𝑑𝑡 (2 − 0)
2 2

𝑑2 (2 − 𝑡)2 (0) − 2(2)(2 − 𝑡)(−1)

[𝑀0 (𝑡)] = 4
𝑑𝑡 2 (2 − 𝑡)

4(2 − 𝑡) 4
= =
(2 − 𝑡)4 (2 − 𝑡)3

′
𝑑2 4 4 1
𝜇2 = 2 [𝑀0 (𝑡)]𝑡=0 = = =
𝑑𝑡 (2 − 0)
3 8 2

1
𝜇1 ′ = 𝑚𝑒𝑎𝑛 =
2

2
𝑉𝑎𝑟𝑖𝑎𝑛𝑐𝑒 𝑖𝑠 𝜇2 = 𝜇2 ′ − 𝜇1 ′

1 1 2
= −( )
2 2

1
=
4

EM III_SMITA N 37