Scribd
Upload
77 views3 pages

Solutions 12

This document contains solutions to problems on a statistics midterm exam. 1) It shows that two random variables X and Y are not independent based on their joint and marginal densities not being equal. It also derives the conditional density of Y given X and finds the expected value of Y given X. 2) It shows that a sequence of random variables {Xn} forms a martingale by taking expectations conditioned on prior values. 3) It finds the probabilities of a binary random variable X being 0 or 1 based on the distribution of a random variable A that X depends on. 4) It performs a variable transformation on two random variables and shows the transformed variables are independent normal random variables based on their joint

Uploaded by

mohamed
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
77 views3 pages

Solutions 12

This document contains solutions to problems on a statistics midterm exam. 1) It shows that two random variables X and Y are not independent based on their joint and marginal densities not being equal. It also derives the conditional density of Y given X and finds the expected value of Y given X. 2) It shows that a sequence of random variables {Xn} forms a martingale by taking expectations conditioned on prior values. 3) It finds the probabilities of a binary random variable X being 0 or 1 based on the distribution of a random variable A that X depends on. 4) It performs a variable transformation on two random variables and shows the transformed variables are independent normal random variables based on their joint

Uploaded by

mohamed
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/ 3

Statistics 351 Fall 2007 Midterm #1 Solutions

1. (a) By definition,

x y

fX (x) =

2e e

dy = 2e

and

Z
fY (y) =



= 2e2x , x > 0,
e

y

y
x y

2e e

dx = 2e


 y

= 2ey 1 ey ,
e

x

y > 0.

1. (b) Since fX,Y (x, y) 6= fX (x) fY (y), we immediately conclude that X and Y are not
independent random variables.
1. (c) By definition,
fY |X=x (y) =

fX,Y (x, y)
2exy
=
= exy , 0 < x < y < .
fX (x)
2e2x

1. (d) By definition,
Z

E(Y |X = x) =

y fY |X=x (y) dy =

y exy dy.

Let u = y x so that du = dy and the integral above becomes


Z
Z
Z
Z
xy
u
u
ye
dy =
(u+x)e du =
ue du+x
eu du = (2)+x(1) = 1+x
x

and so E(Y |X) = 1 + X.


1. (e) Using (d) we find E(Y ) = E(E(Y |X)) = E(1 + X) = 1 + E(X). However,
Z
Z
E(X) =
x fX (x) dx =
2xe2x dx.

Let u = 2x so that du = 2dx and the integral above becomes


Z
Z
1 u
1
1
2x
2xe
dx =
ue du = (2) = .
2 0
2
2
0
Therefore,
E(Y ) = 1 + E(X) = 1 +

1
3
= .
2
2

1. (f ) If U = X + Y and V = X, then solving for X and Y gives


and Y = U V.

X=V
The Jacobian of this transformation

x

u
J =
y

u

is

x

v 0 1
=
= 1.
y 1 1

v

Therefore, we conclude
fU,V (u, v) = fX,Y (v, u v) |J| = 2ev(uv) 1 = 2eu
provided that 0 < 2v < u < (or, equivalently, 0 < v < u2 < ). The marginal for
U is given by
Z
Z u/2
u
fU (u) =
fU,V (u, v) dv =
2eu dv = 2eu = ueu , u > 0.
2

0
We recognize this as the density of a (2, 1) random variable. That is, U = X + Y
(2, 1) as required.
2. (a) We find
E(Xn+1 |X1 , . . . , Xn ) = E(Xn Yn+1 |X1 , . . . , Xn )
= Xn E(Yn+1 |X1 , . . . , Xn ) (by taking out what is known)
= Xn E(Yn+1 ) (since Yn+1 is independent of X1 , . . . , Xn )
= Xn 1
= Xn
and so {Xn , n = 1, 2, . . .} is, in fact, a martingale.
2. (b) For n = 1, 2, . . ., we find
E(Xn ) = E(Y1 Y2 Yn ) = E(Y1 ) E(Y2 ) E(Yn ) = 1
using the fact that Y1 , Y2 , . . . are independent.
3. (a) By the law of total probability,
Z
P (X = 0) =

Z
P (X = 0|A = a)fA (a) da =

and
Z
P (X = 1) =

1
1
a2
(1 a) da = a =
2 0 2

Z
P (X = 1|A = a)fA (a) da =

1
a2
1
a da = = .
2 0 2

3. (b) By definition,
fA|X=0 (a) =

(1 a) 1
P (X = 0|A = a)fA (a)
=
= 2(1 a),
P (X = 0)
1/2

0 < a < 1,

and

P (X = 1|A = a)fA (a)


a1
=
= 2a, 0 < a < 1.
P (X = 1)
1/2

If U = 3 X + Y and V = X 3 Y , then solving for X and Y gives

3U + V
U 3V
X=
and Y =
.
4
4
fA|X=1 (a) =

4.

The Jacobian of this transformation is




x x



u v 3/4

1/4
=
= 1/4.
J =
1/4 3/4
y y


u v
Therefore, we conclude

fU,V (u, v) = fX,Y


=

1
fX
4

!
3u + v u 3v
,
|J|
4
4
!

!
3u + v
u 3v
fY
, < u, v < ,
4
4

using the assumed independence of X and Y . The exact form of fX and fY gives

!2

!2

1
1
1
3u + v
1
1 u 3v
fU,V (u, v) = exp
exp
2

4
4
4
2
2



1
1  2
2
2
2
=
exp
3u + 2 3 uv + v + u 2 3 uv + 3v
8
2 16



1
1
2
2
=
exp
4u + 4v
8
2 16




u2
u2
1
1
exp
= exp
24
24
2 2
2 2
provided that < u < , < v < . Hence, we immediately conclude
that fU,V (u, v) = fU (u) fV (v) and so U and V are independent random variables.
Furthermore, we recognize that both U and V have a N (0, 4) density. Together, this
implies that U |V = v N (0, 4).

You might also like