You are on page 1of 3

UNIVERSITY OF GHANA

(All rights reserved)


DEPARTMENT OF STATISTICS AND ACTUARIAL SCIENCE
STAT 332: PROBLEM SHEET ONE

1. Show that if matrix A > 0 then, A is non-singular and |A| > 0.


 
1 θ
2. If the matrix B = where θ2 < 1, find the matrix Γ and Λ such that B =
θ 1
ΓΛΓT . Hence find B 2 .

3. Let X, Y, Z be independent and uniformly distributed on [0, 1]. Find P (XY < Z 2 ).

4. Let X, Y, Z be independent and uniformly distributed on [0, 1]. Find E([X + Y ]Z 2 )

5. Suppose (X, Y ) has the bivariate Pareto distribution with density function
 12
(x+y−1)5
, x, y > 1
f (x, y) =
0 otherwise.

Find the covariance matrix of (X, Y ).

6. A random vector Z = (Z1 , Z2 )T has dispersion matrix


X  1 ρ 
= , let 0 < p2 < 1.
ρ 1
1 1 1 1
Let 2a = (1 + ρ) 2 + (1 − ρ) 2 and 2b = (1 + ρ) 2 − (1 − ρ) 2 ,
then by using the Spectral
P Decomposition Theorem, solve for the symmetric positive
definite square root of in terms of a and b.

pPn
2
7. If a = (a1 , a2 , ..., an ) , b = (b1 , b2 , ..., bn ) and d(a, b) = i=1 (ai − bj ) is the Euclidean
distance between a and b, and A is an orthogonal matrix of order n, show that
d(a, b) = d(aA, bA)

8. Let (X1 , X2 , X3 ) be a random vector with a multivariate distribution with the expected
value (0, 0, 0) and the variance/covariance matrix:
 
4 1.5 1
Σ =  1.5 1 0.5 
1 0.5 1

If a random variable W is defined by the equation X1 = aX2 + bX3 + W and it is


uncorrelated with X2 and X3 . Calculate the value of the coefficient a.

–KD-A/GK-D, Page 1 of 3
9. Prove that

(a) Cov(X + Y, Z) = Cov(X, Y ) + Cov(Y, Z) where X and Y are two random vectors
of the same dimension and Cov(X, Z) = Cov(Z, X).
(b) If m=n, V (X + Y ) = V (X) + V (Y ) + 2Cov(X, Y )

10. For random vectors X and Y with Cov(X, Y ) = Σ and constant matrices A, B. Prove
the following

(a) Cov(AX, Y ) = AΣ
(b) Cov(X, BY ) = Σ B T
(c) Cov(AX, BY ) = A Σ B T

11. What does it mean to say that a random vector X of dimension k has a) a singular
distribution and b) a non-singular distribution.

12. Independent random variables X = (X1 , X2 )T and Y = (Y1 , Y2 )T have respective mean
vectors and covariance matrices
   
T 3 2 T 4 2
U1 = (2, 2) , Σ1 = , U2 = (3, 2) , Σ2 =
2 4 2 0

What are the mean vector and covariance matrix of

U = (X1 , X2 + Y1 , Y2 )T

13. Let X = (X1 , X2 , X3 )T be N (µ, Σ) where µ = (1, 0, −1)T and


 
1 0 1
Σ= 0 1 1 
1 1 2

(a) Determine the joint distribution of (Y1 , Y2 , Y3 ) where

Y1 = 2X1 − X2 + X3 , Y2 = X1 + 2X2 + X3 , Y3 = −X1 − 2X2 + 2X3 .

(b) What are the marginal distributions of Y1 , Y2 and Y3


(c) Find the correlation matrix of (Y1 , Y2 , Y3 )T

14. If X = (X1 , X2 , X3 , X4 )T has a multivariate Normal distribution N (µ, Σ), with


   
2 3 0 1 4
 3   1 5 0 4 
µ=
 −2  and Σ = 
  .
0 0 2 5 
3 4 4 5 8

–KD-A/GK-D, Page 2 of 3
(a) Determine the joint distribution of (Y1 , Y2 ) where

Y1 = X2 + X3 − 2X4 + 6, Y2 = X1 − 2X3 + X4

(b) Calculate the correlation coefficient of (Y1 , Y2 )

15. A random vector (X, Y ) has the  bivariate


 normal distribution with mean (0, 0) and
1 0
the variance-covariance matrix . Find the probability that (X, Y ) lies in the
0 1
unit circle centered at the origin.

–KD-A/GK-D, Page 3 of 3

You might also like