What is the sufficient statistic for bivariate normal if the covariance matrix is known.
Thanks!
I am puzzled about this. From the definition of sufficient statistics, it seems that 2 df chi-square statistic is the sufficient statistic. But since there are two means, the sufficient statistic should be 2 dimensional?
Last edited by Dason; 04-10-2011 at 07:27 PM.
How exactly do you make that argument?Originally Posted by hehe1223
f(x|theta)=f(t|theta)*h(x)
By factorization theorem.
Posting Permissions
|
|
Reply With Quote