Beruflich Dokumente
Kultur Dokumente
Note: Some of these questions involve more extensive reflection than others and would not
therefore be appropriate for a two hour exam.
1. Suppose that X1 , X2 , are independent random variables with common density function
f (x) = x1
for x 1
where > 0.
(a) Let Vn = max(X1 , , Xn ). Find the density function of Vn . (Be sure to specify the
range of Vn .)
d
(b) Show that n1/ Vn V where the random variable V has distribution function
FV (x) = exp x
for x > 0.
2. Suppose that X1 , X2 , are independent random variables with E(Xi ) = 0 and E(Xi2 ) =
2.
(a) Define
!1/2
!
n
n
1 X
1X
2
X
Xi .
Un =
n i=1 i
n i=1
Find the limiting distribution of Un as n .
(b) Define
n
1X
Vn =
(1)k Xk
n k=1
p
Show that Vn 0.
3. Suppose that X1 , , Xn are independent discrete random variables with mass function
f (x; ) =
!|x|
1
2
!1|x|
for x = 1, 0, 1
n i=1
Let T =
Pn
i=1
n t
P (T = t) =
(1 )nt
t
for t = 0, , n
(a) State why T is sufficient for . (Hint: write the joint probability mass function as a 1
parameter exponential family.)
(b) Find an unbiased estimator of (1 ) based on T when n = 2.
5. Let X and Y be independent random variables with E(X) = E(Y ) = , Var(X) = 12
and Var(Y ) = 22 . Assume that 12 and 22 are known and consider estimators of of the
form
b = aX + (1 a)Y.
b
Find the value of a that minimizes Var().
n
X
1
X2
n(k + 1) i=1 i
n
X
n = 1
Xi .
X
n i=1
1
exp(x/(ti )) for x 0
ti
is an unbiased estimator of .
(b) Show that (b )/ is an approximate pivot for and find an approximate 95% confidence interval for .
8. Suppose that X1 , X2 , are independent uniform random variables on [0, 1]. Suppose
that {bn } is a sequence of estimators with bn > 0 satisfying
n
1X
1
1
=
+
ln(Xi )
n i=1
bn bn + 1
p
(a) Show that bn 0 and find the value of 0 .
n
X
I(xi = 0)
i=1
i:xi >0
2
(a) Show that 2 ni=1 Xi and 2 m
i=1 Yi are independent random variables with 2n and
2m degrees of freedom respectively.
(b) Using the pivot
X
Y
describe in detail how you would construct a 95% confidence interval for /. What is the
distribution of the pivot?
n
X
I(Xi is even) =
i=1
n X
I(Xi = 2k)
i=1 k=0
(where I(A) is 1 if A is true and 0 if A is false). Find an estimator en based on Y such that
p
en .
13. Suppose that Y1 , , Yn are independent discrete random variable with the probability
mass function of Yi given by
yi
for y = 0, 1, 2,
f (y; i ) =
(1 + i )y+1
where i = E(Yi ). Suppose that ln(i ) = 0 + 1 xi where x1 , , xn are known constants.
(a) Show that Var(Yi ) = i (1 + i ).
(b) Show that the maximum likelihood estimators, b0 and b1 , of 0 and 1 satisfy the
equations
)
(
n
n
X
X
exp(b0 + b1 xi )
Yi =
(Yi + 1)
1 + exp(b0 + b1 xi )
i=1
i=1
and
n
X
i=1
xi Yi =
n
X
i=1
xi exp(b0 + b1 xi )
(Yi + 1)
.
1 + exp(b0 + b1 xi )