Intermediate Statistics: Homework 8 Additional Problems - Saddlepoint Approximation - Prof, Assignments of Statistics

Problem a1 from homework 8 of the intermediate theory of statistics course (stat610) taught by jianhua huang. The problem asks students to show that the log-likelihood function k(θ) is a convex function and that the maximum likelihood estimator (mle) θ^ is uniquely defined. Useful for university students taking advanced statistics courses and is most likely associated with the typology of 'exercises' or 'problems'.

Typology: Assignments

Pre 2010

Uploaded on 02/10/2009

koofers-user-dx0
koofers-user-dx0 🇺🇸

10 documents

1 / 1

Toggle sidebar

This page cannot be seen from the preview

Don't miss anything!

bg1
Stat610: Intermediate Theory of Statistics
Homework 8, Additional Problems (10/20/06)
Instructor: Jianhua Huang Editor: Krista Rister
TA: Seokho Lee
Problem A1.
When we derived the saddlepoint approximation for the distribution of the sample mean, we considered the
exponential family
pθ(x) = exp{θx K(θ) + h(x)}
and used the fact that the MLE
b
θis the solution of the “saddlepoint equation” K(
b
θ) = ¯x.
Show that (i) K(θ) is a convex function; (ii) the MLE
b
θis uniquely defined, so there is an 1-to-1 correspon-
dence between
b
θand ¯x.
1

Partial preview of the text

Download Intermediate Statistics: Homework 8 Additional Problems - Saddlepoint Approximation - Prof and more Assignments Statistics in PDF only on Docsity!

Stat610: Intermediate Theory of Statistics

Homework 8, Additional Problems (10/20/06)

Instructor: Jianhua Huang Editor: Krista Rister TA: Seokho Lee

Problem A1.

When we derived the saddlepoint approximation for the distribution of the sample mean, we considered the exponential family pθ(x) = exp{θx − K(θ) + h(x)}

and used the fact that the MLE θ̂ is the solution of the “saddlepoint equation” K′( θ̂) = ¯x.

Show that (i) K(θ) is a convex function; (ii) the MLE ̂θ is uniquely defined, so there is an 1-to-1 correspon-

dence between θ̂ and ¯x.