Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
TAMS32/TEN1 STOKASTISKA PROCESSER TENTAMEN LÖRDAG 7 JANUARI 2017 KL 08.00-12.00. Examinator och jourhavande lärare: Torkel Erhardsson, tel. 28 14 78. Permitted exam aids: Formel–och tabellsamling i TAMS32 Stokastiska processer (handed out during the exam). Mathematics Handbook for Science and Engineering (formerly BETA), by L. Råde och B. Westergren. Calculator with empty memories. The exam consists of 6 problems worth 3 points each. Grading limits : 8 points for grade 3, 11.5 points for grade 4, 15 points for grade 5. The results will be communicated by email. Uppgift 1 Let {Xt ; t = 0, 1, . . .} be a Markov chain with state space SX = {0, 1, 2}, initial distribution p(0) and transition matrix P , where 1/3 a 2/3 1/2 b . P = 1/2 0 p(0) = 1/2 , c 3/4 0 0 (a) Compute the constants a, b and c. (b) Compute p(2), that is, compute the probabilities P (X2 = i), i = 0, 1, 2. (c) Does the chain have a stationary distribution? Does it have an asymptotic distribution? Motivate your answers. Compute these distributions if they exist. Uppgift 2 Let {B(t); t ≥ 0} be a Wiener process (Brownian motion) with variance parameter σ 2 = 1. The Brownian bridge process {X(t); 0 ≤ t ≤ 1} is defined by: X(t) = B(t) − tB(1) ∀0 ≤ t ≤ 1. (a) Compute P (−1 < X( 21 ) < 1). (b) Let Y = X( 21 ) − X(0) and Z = X(1) − X( 12 ). Compute the correlation coefficient ρ(Y, Z). (c) Does the Brownian bridge have independent increments? Motivate your answer. Uppgift 3 Let {X(t); t ∈ R} be a white noise process in continuous time, with spectral density SX (f ) = 1 ∀f ∈ R. Decide whether a wide sense stationary process {Y (t); t ∈ R} with mean µY = 0, and autocovariance function CY (τ ) = e−2|τ | ∀τ ∈ R, can be obtained as the output signal of a linear time invariant filter (LTI), when the input signal is the white noise {X(t); t ∈ R}. Motivate your answer. If the answer is yes, give the impulse response for such an LTI. Uppgift 4 Let {Yt ; t ∈ Z} be the wide sense stationary AR(2) process which solves the equation 1 1 Yt − Yt−1 − Yt−2 = Xt ∀t ∈ Z, 6 6 where {Xt ; t ∈ Z} is white noise with E(Xt ) = 0 and V (Xt ) = 1. The autocorrelation function of {Yt ; t ∈ Z} is RY (τ ) = 48 1 |τ | 27 1 |τ | ( ) + (− ) 70 2 70 3 ∀τ ∈ Z. (You do not have to show this.) (a) Compute the LMMSE of Yt based on (Yt−1 , Yt−2 ), and the corresponding mean square prediction error. (b) Compute the LMMSE of Yt based on Yt−1 , and the corresponding mean square prediction error. Uppgift 5 Let X and Y be√independent random variables, such that X ∼ N(0, 1), and √ Y ∼ 3, 3), that is, Y is uniformly distributed over the interval √ Uniform(− √ (− 3, 3). Let Z(t) = X cos(ωt) + Y sin(ωt) ∀t ∈ R, where ω > 0 is a constant. Is the stochastic process {Z(t); t ∈ R} wide sense stationary? Is it stationary? Motivate your answer(s). Hint: It might be useful to know that cos(α+β) = cos(α) cos(β)−sin(α) sin(β), and that cos(α − β) = cos(α) cos(β) + sin(α) sin(β). Uppgift 6 A bus starts from its terminus at time T0 = 0, with 10 passengers. The bus lets off one passenger at a time, at times 0 < T1 < T2 < · · · < T10 , which are given by a Poisson process {N (t); t ≥ 0} with intensity 1. (That is: 0 < T1 < T2 < · · · < T10 are the first 10 jump times of the Poisson process.) Let Xi (i = 1, . . . , 10) be the time it takes for passenger no. i to reach his/her home after he/she has disembarked from the bus. Assume that the random variables X1 , X2 , . . . , X10 are independent and Exponential(λ) distributed, and also independent of the Poisson process. Compute the probability that none of the 9 first passengers have reached their homes at time T10 (when the last passenger is let off from the bus). Compute also the numerical value of this probability when λ = 0.1. Hint: Compute first the conditional probability that none of the passengers have reached their homes at time T10 , given the event {T1 = t1 , T2 = t2 , . . . , T10 = t10 }. Then, use some suitable properties of the random variables T1 , T2 , . . . , T10 .