Digital Signal Processing Reference
In-Depth Information
The second moment is equal to:
1
x
2
l
e
lx
d
x ¼ l
Gð
3
Þ
l
3
2
2
l
2
:
!
l
2
¼
m
2
¼
¼
(2.378)
0
From (
2.377
) and (
2.378
), using (
2.337
), the variance is:
2
l
2
1
l
2
¼
1
l
2
:
s
2
¼ m
2
m
1
¼
(2.379)
The third central moment is:
1
3
l
e
lx
d
x:
m
3
¼
ðx
1
=lÞ
(2.380)
0
After brief calculations, we arrive at:
=l
3
m
3
¼
2
:
(2.381)
The nonzero third central moment is the result of asymmetry in the PDF.
Similarly, we get the fourth central moment:
1
4
=l
4
l
e
lx
d
x ¼
9
m
4
¼
ðx
1
=lÞ
:
(2.382)
0
From (
2.372
) and (
2.379
)-(
2.381
), the coefficient of skewness is given by:
3
m
3
s
3
¼
EfðX mÞ
g
c
s
¼
¼
2
:
(2.383)
s
3
The positive value (
2.383
) shows that the density function is skewed to the right,
as it is shown in Fig.
2.51
.
Similarly, from (
2.373
), (
2.379
), and (
2.382
), the coefficients of kurtosis is
given as:
4
m
4
s
4
¼
EfðX mÞ
g
c
k
¼
¼
9
:
(2.384)
s
4
The large coefficient of Kurtosis indicates that the density has a sharp peak near
its mean. This is confirmed in Fig.
2.51
.
Higher values of the moments do not have a similar physical interpretation, as do
the first four moments.
Search WWH ::
Custom Search