Scribd
Upload
139 views14 pages

IDL 5 - Moment and Moment Generating Function

The random variable x has an exponential distribution with parameter λ = 2. The moment generating function of an exponential distribution is: Mx(t) = E(etx) = ∫_0^∞ ketx e^-λx dx = k/ (λ - t) = 1/(2 - t) (ii) y = 2 + x Using the property that if y = a + bx, then My(t) = eata Mx(bt), we have: My(t) = e2t Mx(2t) = e2t / (2 - 2t) = e2t

Uploaded by

Reagan Torbi
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
139 views14 pages

IDL 5 - Moment and Moment Generating Function

The random variable x has an exponential distribution with parameter λ = 2. The moment generating function of an exponential distribution is: Mx(t) = E(etx) = ∫_0^∞ ketx e^-λx dx = k/ (λ - t) = 1/(2 - t) (ii) y = 2 + x Using the property that if y = a + bx, then My(t) = eata Mx(bt), we have: My(t) = e2t Mx(2t) = e2t / (2 - 2t) = e2t

Uploaded by

Reagan Torbi
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 14

Moment and Moment Generating

Function
Moments

Let x be the random variable with probability distribution,


function f(x) and g(x) be real-valued function of x. Then

E[g(x)]= 𝑥 𝑔 𝑥 𝑓 𝑥 , if x is discrete

= 𝑥
𝑔 𝑥 𝑓 𝑥 , if x is continuous
Moment and Moment Generating Function
The kth Moment about the origin

If g(x)  x , we obtain the kth moment about the origin, denoted Uk.
k

and defined by
U ki = E (xk) =  x f (x) or 
x
k Rx xk f (x) dx , where

U11 = E(x) = x x f (x) or  Rx x f (x)dx ,

which is the mean and also called the first moment about
origin
U12 =
x x f (x) or  x f (x) dx ,
2 Rx 2

which is called the second moment about the


origin.
Moment and Moment Generating Function
The kth Moment about the mean

If g(x)  (x  )k , we get the kth moment about the mean, denoted
and defined by
Uk = E[( x  k ] =  (x  )k f (x) or R (x  )k f (x) dx
) x x
Uses of moments about the mean in statistical analysis

• 𝑼𝟐 = 𝑬(𝒙 − 𝝁)𝟐 the second moment about the mean


also known as variance.

• 𝑼𝟑 = 𝑬(𝒙 − 𝝁)𝟑, the third moment about the mean


describes
the skewness 𝑈3 of a distribution. The measure of
given by 𝑎3 = .if a3  0, the distribution becomes skewed
skewness is 𝜎 3

(that is, tailed to the right or left depending on whether a3 >


0 or a3 < 0)
Uses of moments about the mean in statistical analysis

• 𝑼𝟒 = 𝑬(𝒙 − 𝝁)𝟒 the fourth moment about the mean is the


peakness (or kurtosis) of a distribution. The degree of

peakness is 𝑎4 = 𝑈 4.
𝜎4

• If a4 = 3, the distribution is normally distributed.

• If a4 < 3, the distribution flattens at the centre than the

normal distribution.
• If a4 > 3, the distribution becomes more peaked at the

centre than the normal distribution


Expansion of the moments about the mean

(i) The second moment about the mean,

U2
= E[( x  )2 ]

= E[x  2 x   ]=
2 2
E(x 2
)
(ii)
2
The third moment about the mean,
U = E[( x  )]
3
3

= E[x3  3x2  3 2
3
Expansion of the moments about the mean
(iii) The fourth moment about the mean,
U = E[( x  ) ]
4
4

= E[x  4x  6 x  4  x   4]
4 3 2 2 3

= E(x 4 )  4E(x3 )  6 2 E(x 2 )


 3 4
Moment Generating Function

Moments of most distributions can also be determined by


finding a function in a form of series. The coefficients of the
series give the moments. The function which generates the
moments is called moment generating function. If it exists, the
mgf for the distribution function, f(x) is given by:

= E(e ) = e Rx f
tx
M x (t) f (x) e
or tx
(x)dx
tx

x
Expansion of 𝒆𝒕𝒙

Now expanding the function, etx and taking


expectation, 2 2 3 k k

etx = 1 + tx + t x  t x + . . . + t x
3

2! k
3! !
M x (t) = E(etx) = E(1 + tx t 2 x 2  t3 + . . . + t k xk
2! k
+ 3
x 3! ) !
= 1 + t E(x) t 2 E(x2) t 3 E(x3) +. . . tkE k

+ 2! + 3! + k! (x )
= 1 t1 U1  t2 U1  t3 U1  . . . tk U1
1 2 3 k
1
! 2! 3! k
!
The coefficient tk is Uk1 , the kth moment about the origin,
of k! which
is also obtained by taking the kth derivative of M (t) with
x

respective to t and evaluating it at t = 0. That is,


k k Mx(t)
E (x ) t = Uk1i = M x(0)
tk 0
=
Application of L’Hospital Rule

h(t)
In evaluating M 1x (t) which takes the form M x1 (t) 
q(t ) at t  0 w e m a y

obtain the indeterminate, 0


0 or 
 . In such cases we apply the
L’Hopital Rule, where
h1 (t)
lim M (t)  lim t
t 0
1
x
t  0 q (t)
.

For example, if x is uniformly distributed over the interval [0,


1
1], then its moment generating function, M x (t)
t
= (e t  1) .
Differentiating w ith respect to t a n d evaluating at t = 0;
tet  et  1  h(t)
Mx 1
(t) 
t 2 q (t )

M 1
(0)  0
x
0 ,

w h i c h i s i n d e t e r m i n a t e A a p p l y i n g t h e L’ H o p i t a l R u l e w e h a v e ,
h1 ( x )
lim M
t 0
1
x (t) = lim 1
t  0 q (x)

tet  et  et
= lim
t 0 2t

tet 1 t 1
= lim
t  0 2t
= li m e  ,
t 0 2 2

w h i c h is the m e a n v a lu e of the u n i f o r m distribution.


Properties of Moment Generating Functions

• Moment generating functions are unique.


• If x and y are random variables such that y = a + bx, then
M y (t) = E(e )
ty

= E (e ( a bx)t )

= E(e a t .ebtx )

= eat E (ebtx )
= x
e a t .M (bt )
Ex 1:

Determinethe moment generating functions forthe


random variables, x and y with the following distribution
functions:
(i
)  kxe2 x
, x
f (x) 
 0 0 , elsewhere

You might also like