Scribd
Upload
9 views1 page

Statistical Inference: Likelihood Ratio Tests

The document discusses statistical inference and likelihood ratio tests. It provides 4 examples of deriving likelihood ratio tests for different distributions including Bernoulli, Pareto, and exponential distributions. The tests compare null hypotheses about distribution parameters to alternative hypotheses.

Uploaded by

hibezowo
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
9 views1 page

Statistical Inference: Likelihood Ratio Tests

The document discusses statistical inference and likelihood ratio tests. It provides 4 examples of deriving likelihood ratio tests for different distributions including Bernoulli, Pareto, and exponential distributions. The tests compare null hypotheses about distribution parameters to alternative hypotheses.

Uploaded by

hibezowo
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

SQB7024: STATISTICAL INFERENCE

Tutorial 4

1. Let X 1 ,....X n be a random sample from a population N  , 2 . Consider 


testing H 0 :  ≤ θ 0 versus H1 :  >  0 , where  >  0 is a fixed positive
number. Assuming that  is unknown, derive the likelihood ratio test.

2. Suppose that we observe m iid Bernoulli() random variables, denoted by


Y1 ,Y2 ,...,Ym . Show that the LRT of H 0 :  ≤ θ 0 versus H1 :  >  0 , will


m
reject H 0 if Y  b.
i 1 i

3. A random sample X1,.... X n is drawn from a Pareto population with pdf

 
f  x |  ,  = x ,   0,   0
 1  , 
I
x

(a) Find the MLEs of  and  .


(b) Show that the LRT of

H 0 :  = 1,  unknown versus H1 :   1,  unknown

has critical region of the form x : T  x   c1 or T  x   c2  , where


0  c1  c2 and

 n 
  Xi 
T  x   log  i 1 
 
n
  min X i  
 i  

4. Suppose that we have two independent random samples: X1,.... X n are


exponential () and Y1 ,....Yn are exponential().

(a) Find the LRT of H 0 :  = , versus H1 :   .


(b) Show that the test in part (a) can be based on the statistics

 Xi
T x 
 X i   Yi

You might also like