Unlock access to all the studying documents.
View Full Document
Chapter 16
Bayesian Estimation and Inference
◼ Exercises
1 (a) The likelihood function is
The product of factorials will fall out. This leaves
1
0
exp( ) (1 / )
( | , , )
exp( ) (1 / )
ii
ii
y
ny
n
p y y
nd
−
=−
Application
a. p(Fi|Ki,) =
−
−
(1 )
i i i
iF K F
i
K
F
so the log-likelihood function is
11
1
()
(1 ) (1 )
( ) ( ) .
i i i
niF K F ab
i
i
Kab
Fab
−−−
=
+
− −
This simplifies considerably. The combinatorials and gamma functions fall out, leaving
11 () 11
1
(1 ) (1 ) (1 ) (1 )
i i i i i i i i
nF K F ab F K F ab
i
−−− − −−
=
− − − −
The denominator is a beta integral, so the posterior density is
The denominator simplifies slightly;
( ) ( 1) [ ( )] ( 1)
[( ) ( 1)] [( ( )) ( 1)]
( | ) (1 )
[( ) ( 1) ( 1)]
i i i i i
i i i i i F a K F b
ii
F a K F b
pK a b
+ − − + −
+ − − + −
= −
+ − + −
y