Suppose the cdf of (X1 , X2) is given by
F(x1,x2) =1-e-x1 + e-x2 + e-x1-x2
for x1 ≥ 0,x2 ≥ 0 and is 0 otherwise.
a)(5 marks) Determine the joint probability density function of (X1 , X2).
b)(5 marks) Are and statistically independent? Justify your answer.
c)(5 marks) Determine the mean vector and variance matrix of (X1 , X2).
d)(5marks) Determine the mean vector and variance matrix of where
e)(5 marks) Are and in (d) statistically independent? Justify your answer.
f)(5 marks) Determine the joint density function of (Y1 , Y2)
Suppose Z0 , Z1,…are i.i.d.N(0,1). With T = {1,2,…}, define the process by {(t,Xt) : t∈T } by Xt = ZtZt-1
a) (5 marks) Determine the mean and autocovariance functions of the Xt process.
b) (5marks)If Z0=1 , Z1=3. , Z2=-4, Z3=2 then plot the first three values of the sample function of the Xt process.
(c) (10 marks) Determine the moment generating function (mgf) mXt(s) of Xt (Hint: use the theorem of total expectation.) Does the mgf exist for all s ∈R1
Suppose that h : R4 → R1 is given by
(a) (5 marks) Prove that is h convex.
(b) (5 marks) Suppose that X = (X1, X2, X3, X4)’ has a joint distribution with mean vector and variance matrix given by
Determine a general lower bound on E(h(X)) .
(c) (5marks) If X~N4(μ,∑) , then determine E(h(X)) exactly.
(d) (5 marks) What is the best afinepredictor of when X2 =(X3, X4) when X1 = (X1, X2)=(x1,x2) is observed? Under what conditions is this also the best predictor and explain what ”best”
(10 marks) Suppose that Z0,Z1,… are i.i.d N(0,1) and Xn = 1/n + αZn + βZn-1 . Determine whether or not {(n,Xn ):n∈ N } is a stationary Gaussian process.