| |
 |
|
|
Science Forum Index » Statistics - Math Forum » Question on moment generating function
Page 1 of 1
|
| Author |
Message |
| Guest |
Posted: Sun Apr 13, 2008 2:24 am |
|
|
|
|
Hi,
I came across this example in one of the books that I'm reading.
Given X~Exp(a), then for u<a
mgf x(a) = a * integral from 0 to infinity (exp(ux)exp(-ax)dx = a/(a-
u)
I know that mgf x(a) = integral from -infinity to infinity
(exp(ux)dFx(x)), however, how I could substitute the dFx(x) term with
the given X term to obtain the results given above? Sorry, my stats is
a bit scratchy.
Thanks. |
|
|
| Back to top |
|
| Paul Rubin |
Posted: Sun Apr 13, 2008 10:11 am |
|
|
|
Guest
|
thampw@hotmail.com wrote:
Quote: Hi,
I came across this example in one of the books that I'm reading.
Given X~Exp(a), then for u<a
mgf x(a) = a * integral from 0 to infinity (exp(ux)exp(-ax)dx = a/(a-
u)
I know that mgf x(a) = integral from -infinity to infinity
(exp(ux)dFx(x)), however, how I could substitute the dFx(x) term with
the given X term to obtain the results given above? Sorry, my stats is
a bit scratchy.
Thanks.
Since X ~ Exp(a), for x >= 0 F(x) = 1 - exp(-ax) and dF(x) = f'(x)dx =
a*exp(-ax)*dx. For x < 0, F(x) = 0 and dF(x) = 0; so the lower limit of
integration changes from -infinity to 0.
Note the restriction u < a on the formula for the MGF.
/Paul |
|
|
| Back to top |
|
| Guest |
Posted: Tue Apr 15, 2008 10:46 pm |
|
|
|
|
Thanks for the explanation, Paul.
Cheers! |
|
|
| Back to top |
|
| |
|
Page 1 of 1
All times are GMT - 5 Hours
The time now is Fri Aug 29, 2008 8:29 pm
|
|