TOPIC 6 (Part 1)

MOMENTS &
MOMENT GENERATING FUNCTION

1
PROBABILITY AND STATISTICS | CENTRE OF STATISTICS AND DECISION SCIENCE
 By the end of this lesson, you should be able to:

Learning ▪ Calculate expectation and variance by

Objectives
calculating the moments for discrete and
random variables.
▪ Compute moments of a random variable by
differentiating the moment generating
function of the random variable.

UiTM MOOC | PROBABILITY AND STATISTICS | TOPIC 6 2

 01 Moment

Subtopics 02 Moment Generating Function

03 Moment Generating Function for

Discrete Random Variable

UiTM MOOC | PROBABILITY AND STATISTICS | TOPIC 6 3

Moment-Generating Functions (MGF)

▪ Moment generating functions (MGF) are a very powerful

computational tool. They make certain computations much
shorter.
▪ It is sometimes more useful to work with the moments of a
distribution instead of the mathematical formula.

UiTM MOOC | PROBABILITY AND STATISTICS | TOPIC 6 4

Moments
Definition 1: Moment
The rth moment about the origin of the random variable X is given by
 x r f ( x), if X is discrete

 x
r = E ( X ) =  
' r

  x r f ( x) dx, if X is continuous

− 

• First moment about the origin: 1 = E ( X )

• Second moment about the origin: 2 = E ( X )
2

• Third moment about the origin: 3 = E ( X 3 ) and so on.

 Mean and variance of a random variable are
 = E ( X ) = 1' UiTMand  2 =ANDVSTATISTICS
MOOC | PROBABILITY =
( X )| TOPIC  − (1' ) 2
6 2 5
Moment Generating Function

Definition 2: Moment Generating Function

The moment-generating function (MGF) of the random variable X is
given by E(etX) and denoted by Mx(t). Hence,


 
e tx f ( x), if X is discrete

 x
M X (t ) = E (e ) =  
tX



− 

 e tx f ( x) dx, if X is continuous

Note: Moment-generating functions will exist only if the

sum or integral of Definition 2 converges.
UiTM MOOC | PROBABILITY AND STATISTICS | TOPIC 6 6
How to get moments from MGF?

Theorem 1:
Let X be random variable with MGF Mx(t). Then

d r M x (t )
r
= E ( X r
) =  r
dt t =0

When r = 1,  = E ( X ) =
' dM x (t )
= M X (0)
' V ( X ) = E ( X 2 ) − [ E ( X )]2
1
dt t =0 =  2' − ( 1' ) 2
= M X (0) − [ M X (0)]2
'' '
2
d M x (t )
When r = 2, 2' = E ( X 2 ) = =
''
2
M X (0)
dt t =0

UiTM MOOC | PROBABILITY AND STATISTICS | TOPIC 6 7

A. Moment Generating Functions
(Discrete Random Variable)

UiTM MOOC | PROBABILITY AND STATISTICS | TOPIC 6 8

Example 1

A random variable X has the discrete uniform distribution

1
 , x = 1, 2, ..., k
f ( x) =  k
0 , elsewhere

Show that the moment-generating function of X is

et (1 − e kt )
M X (t ) =
k (1 − et )

UiTM MOOC | PROBABILITY AND STATISTICS | TOPIC 6 9

Example 2

Suppose that X is a random variable that has the probability function

0.5 , x=6

P( X = x ) = 0.3 , x =8
0.2 x = 10
 ,

Find the moment generating function for X.

UiTM MOOC | PROBABILITY AND STATISTICS | TOPIC 6 10

Example 2 (Solution)

M X (t ) = E (e )
tx

=  e f ( x)
tx

= 0.5 e + 0.3 e + 0.2 e

6t 8t 10 t

UiTM MOOC | PROBABILITY AND STATISTICS | TOPIC 6 11

Example 3

A random variable X has the Poisson distribution for x = 0, 1, 2, …

Show that the moment-generating function of X is

 ( et −1)
M X (t ) = e
Using MX(t), find the mean and variance of the Poisson distribution.

UiTM MOOC | PROBABILITY AND STATISTICS | TOPIC 6 12

Example 3 (Solution)

MGF of X:
− x
e
 
(  e t x
)
M X (t ) = E[e ] =  e
tX tx
=e 
−
=e  ( et −1)

x =0 x! x =0 x!
First moment:
dM X (t ) t  ( e t −1)
 = E[ X ] =
'
1 = e e =

M
dt t =0 t =0

Second moment:
2
d M X (t ) t + et −
 2 = E[ X ] =
' 2
2
=  (1 + e )e
t
=  (1 +  )
dt t =0
t =0

Var[ X ] = E[ X 2 ] − E 2 [ X ] =  (1 +  ) − 2 = 
UiTM MOOC | PROBABILITY AND STATISTICS | TOPIC 6 13
Reference

Walpole, R. E., Myers, R. H., Myers, S. L. and

Ye, K., Probability and Statistics for Engineers
and Scientists, 9th,Global Edition, Prentice
Hall, 2016, ISBN-13: 9780321629111

UiTM MOOC | PROBABILITY AND STATISTICS | TOPIC 6 14

 SANIZAH AHMAD, Senior Lecturer
[email protected]

THANK YOU NORSHAHIDA SHAADAN (DR.), Senior Lecturer

[email protected]

We hope you enjoy

Presenter
ZURAIDA KHAIRUDIN, Senior Lecturer
this lesson. TH [email protected]

If there is any MUHAMMAD ASMU’I ABDUL RAHIM, Lecturer

question, you may

[email protected]

contact any of us. SHAMSIAH SAPRI, Senior Lecturer

[email protected]

15
PROBABILITY AND STATISTICS | CENTRE OF STATISTICS AND DECISION SCIENCE