Digital Signal Processing Reference
In-Depth Information
From (
5.14
), taking
n ¼
1 and
n ¼
2, we obtain the mean and the mean squared
values:
m
1
¼ m
X
¼ X ¼
e
m
Y
þ s
Y
=
2
;
(5.15)
¼
e
2
m
Y
þ
2
s
Y
m
2
¼ X
2
:
(5.16)
The corresponding variance is obtained from (
5.15
) and (
5.16
):
X
¼ m
2
m
1
¼
e
2
m
Y
þ s
2
Y
ð
e
s
2
s
2
Y
1
Þ:
(5.17)
5.1.4 What Does a Lognormal Variable Tell Us?
Consider a large number of independent random variables
Y
1
;
...
; Y
N
;
(5.18)
and the random variable
X
, which is equal to the product of the variables (
5.18
)
X ¼
Y
N
Y
i
:
(5.19)
i¼
1
Taking the natural logarithm of both sides of (
5.19
), we get:
ln
X ¼
X
N
ln
Y
i
:
(5.20)
i¼
1
According to the central
limit
theorem,
the right side of (
5.20
) can be
approximated as a normal random variable
Y
,
Y
X
N
ln
Y
i
:
(5.21)
i¼
1
Taking into account (
5.21
), the expression (
5.20
) becomes:
ln
X Y:
(5.22)
This expression can be rewritten as:
X
e
Y
:
(5.23)
Search WWH ::
Custom Search