Scribd
Upload
46 views2 pages

Homework 3few

This document outlines the homework assignments for STATS 200 taught at Stanford University in Summer 2015. It includes 5 problems analyzing Bayesian statistical concepts such as finding the posterior distribution and mean given certain prior distributions and observed data. The problems cover topics like conjugate prior distributions, uniform distributions, and applying Bayesian methods to problems from a previous homework.

Uploaded by

Jung Yoon Song
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
46 views2 pages

Homework 3few

This document outlines the homework assignments for STATS 200 taught at Stanford University in Summer 2015. It includes 5 problems analyzing Bayesian statistical concepts such as finding the posterior distribution and mean given certain prior distributions and observed data. The problems cover topics like conjugate prior distributions, uniform distributions, and applying Bayesian methods to problems from a previous homework.

Uploaded by

Jung Yoon Song
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

STATS 200 (Stanford University, Summer 2015)

Homework 3:

Due at 1:00 p.m. on July 17

DeGroot & Schervish X.Y .Z means Exercise Z at the end of Section X.Y in our text,
Probability and Statistics (Fourth Edition) by Morris H. DeGroot and Mark J. Schervish.
1. DeGroot & Schervish 7.3.12. Also find the posterior mean and the posterior mode.
2. DeGroot & Schervish 7.4.6.
3. Let X1 , . . . , Xn be iid random variables with a continuous uniform distribution on [0, ],
where > 0 is unknown. Suppose we assign to the prior pdf

q kq

q+1
() =

if k,
if < k,

where k > 0 and q > 0 are constants.


Note: This distribution is called the Pareto(k, q) distribution, and its mean is kq/(q 1)
if q > 1 (and if q 1). You may use these facts without proof.
(a) Find the posterior distribution of .
(b) Find (or simply state) the posterior mean of .
4. Let X1 , . . . , Xn be iid random variables with pdf or pmf of the form
f (x) = f (x ) = exp[() r(x) ()] h(x),
where is unknown, and where , r, , and h are real-valued functions. Suppose we
assign to the prior pdf
() = c exp[a () + b ()],
where a and b are constants, and where c1 = exp[a () + b ()] d < , i.e., c is
simply some appropriate normalization constant. Show that the posterior pdf of has
the form
( x1 . . . , xn ) = c exp[
a () + b ()]
where a
and b do not depend on , and where c1 = exp[
a () + b ()] d < .
Note: Your answer should clearly show how a
and b can be written in terms of some or
all of the following: the constants a and b; the sample size n; the functions , r, , and h;
and the data x1 , . . . , xn .

Homework 3: Due at 1:00 p.m. on July 17

5. We now consider a Bayesian approach to problem 7 of Homework 2. Suppose that is an


unknown positive integer, and suppose that X has the discrete uniform distribution
on {1, . . . , }. Let the prior pmf of be

6 1

2 2
() =

if {1, 2, . . .},
otherwise,

where the that appears as 2 in the denominator is (in this case) the usual mathematical
constant 3.14. Suppose we observe X = x = 715.
Note: To streamline the notation for this problem, define the constants
1
1.4 103 ,
2
k
k=715

a715 =

1
9.8 107 .
3
k
k=715

b715 =

You may wish to use one or both of these constants in your solution.
(a) Find the posterior distribution of .
(b) Find the posterior mean of . (Give an approximate numerical answer based on the
approximate numerical values for a715 and b715 provided above.)

You might also like