MATH 452/STAT 552 Final Exam Supplementary Questions December 4, 2020 The questions below are only based on the materials after the midterm, but the final exam is comprehensive. Question 1 (Continuous-time Markov chains). Consider the Q-matrix Q = a b c a −4 2 2 b 3 −6 3 c 4 4 −8 . (1) Find the parameters λi of the holding times of all states and the transition probability matrix (Pij) of the embedded Markov chain (Pii = 0 for all i). (2) Find the stationary distribution of the embedded Markov chain from (1). (3) Is the probability vector pi = ( 6 13 , 4 13 , 3 13 ) a stationary distribution of the continuous- time Markov chain with Q-matrix given above? Justify your answer. Question 2 (Poisson processes). Let e1, e2 be independent exponential random variables such that E[ei] = λ−1i . Find E[e2|e1 < e2] Question 3 (Renewal processes). Suppose that the interarrival distribution of a renewal process (Nt; t ≥ 0) is the continuous uniform distribution on [0, 2]. (1) Show that the renewal functionm(t) = E[N(t)] satisfiesm(t) = t∧2 2 +1 2 ∫ t∧2 0 m(s)ds, where a ∧ b = min{a, b}. (2) Find limt→∞Nt/t. Question 4 (Brownian motion and Gaussian stochastic integrals). Let B = (Bt; t ≥ 0) be a standard Brownian motion. (1) Find E[B22B21 ]. (Recall that if Z ∼ N (0, σ2), then E[Z4] = 3σ4.) (2) Let It = I0 + ∫ t 0 rdBr, where I0 ∼ N (0, 2) is independent of B. Find E[It] and Var(It). Question 5 (The strong law of large numbers and the central limit theo- rem). Let X1, X2, · · · , Xn, · · · be i.i.d. exponential random variables with mean 2. Write out the following limits explicitly for the sums of independent random variables Sn = ∑n j=1Xj: (1) The strong law of large numbers. (2) The central limit theorem. 1 Answers Question 1. (1) λa = 4, λb = 6, λc = 8. P = a b c a 0 1/2 1/2 b 1/2 0 1/2 c 1/2 1/2 0 . (2) The embedded Markov chain is the random walk on a graph. The graph is the discrete cycle of 3 vertices where each vertex has 2 neighbors. Hence, the stationary distribution is pi(a) = pi(b) = pi(c) = 2/6. (3) Yes. piQ = 0. Question 2. 1 λ1+λ2 + 1 λ2 . Question 3. (1) The required equation is the renewal equation with f(t) = 1 2 1[0,2](t). (2) 1. Question 4. (1) 4. (2) E[It] = 0, Var(It) = 2 + 13t 3. Question 5. (1) P(limn→∞ Sn/n = 2) = 1. (2) limn→∞ P(Sn−2n√4n ≤ x) = 1√2pi ∫ x −∞ e −t2/2dt for all x ∈ R. 2
欢迎咨询51作业君