i*****g
发帖数: 5 | 1 大虾救命！！！
How to find out the MLE of logistic distri.
f(x)=e^-1*(1+e^-1)^-2
it should be related to fisher information. but I do not
know how to calculate
the fisher information number.
also how to prove its variance=pi^2/3???
Thanks | i*****g
发帖数: 5 | 2 sorry, typo, should be f(x)=e^-x*(1+e^-x)^-2 | c******e
发帖数: 18 | 3 To consider MLE of a distribution, in the form of density,
there should have a parameter, so I think maybe u forgot to
include it. Is the density like this:
f(x,a)=e^(-(x-a))/[1+e^(-(x-a))]^2, it's the logistic
distribution.
For this distribution, consider L(x,a), which is
L(x,a)=f(x1,a)*f(x2,a)*...*f(xn,a) and take log both sides,
l(x,a)=log[f(x1,a)]+log[f(x2,a)]+...+log[f(xn,a)]
E{derivative of log[f(X,a)]}=Integral{ [derivative of
f(x,a)/f(x,a)]*f(x,a)dx
=Integral{
【在 i*****g 的大作中提到】
: sorry, typo, should be f(x)=e^-x*(1+e^-x)^-2
|
|
