| |
|
Science Forum Index » Statistics - Math Forum » Bayesian statistics and sufficient statistics...
Page 1 of 1
|
| Author |
Message |
| ... |
 Posted: Wed Jul 02, 2008 4:55 am |
|
|
|
Guest
|
Hi everyone,
I am working on the following question: Suppose that T is a sufficient
statistic for unknown parameter theta, based on a sample X.
a)Prove that the Bayes estimator db(X) of theta is a function of T.
b)Under what conditions can you construct a UMVUE estimate of theta
which is a function of db(X).
The following is my work so far.
a)The posterior distribution theta|X is the same as theta|T since
theta|X
is proportional to
X|theta*theta
=
X|T,theta*T|theta*theta
is proportional to
T|theta*theta
where the last line uses the sufficiency of T
b)We know from a) that db(X) of theta is a function of T. We also need
that db(X) is complete.
Am I right? What can I do to improve my arguments?
I appreciate your help. |
|
|
| Back to top |
|
| |
|
Page 1 of 1
All times are GMT - 5 Hours
The time now is Sat Aug 30, 2008 10:08 am
|
|
