STAT 150: Stochastic Processes (Fall 2020) Test #1 Instructions: • Upload to Gradescope by Thurs, Oct 15, 1:00 PM PST. • Late tests will not be accepted. Please plan accordingly. • Write each question on a separate page. • Write clearly and fully justify all of your work. • Upload an honesty declaration statement. You may type your solutions, but the honesty declaration must be written in your own handwriting. • DSP: If you are 150% time, email me your test by Fri, Oct 16, 1:00AM PST. If you are 200% time, email me your test by Fri, Oct 16, 1:00PM PST. Please email me to discuss any other accommodations that you require. Academic Honesty: Along with your solutions, upload an honesty declaration statement. Your assignment will not be accepted otherwise. In your own handwriting, write out all of the following word for word: • As a member of the UC Berkeley community, I act with honesty, integrity, and respect for others. • I will not communicate with anyone about the exam, besides the instructor and GSI for the entire duration of the exam period. • I will not refer to any books, notes, or online sources of information while taking the exam, other than the course textbooks, lecture notes and workshop materials. • I understand that cheating and plagiarism are serious academic offenses, and will be treated as such, as discussed in the course syllabus. • Sign your name. 1 1. Consider a Markov chain (Xn,n ≥ 0) on a finite state space S, where each state i ∈ S is given a positive weight wi > 0. The Markov chain starts from a uniformly random state X0 in S. At each subsequent step, the Markov chain transitions according to the following two-step procedure. If Xn−1 = i, then • first a state j 6= i is selected uniformly at random, and then, • given this j, with probability w j/(wi+w j) the Markov chain transitions to j (i.e., Xn = j), and with the remaining probability wi/(wi+w j) it stays put (i.e., Xn = i). Find the long run proportion of time spent in any given state i ∈ S. Fully justify your answer. 2. Let α ∈ (0,1). Consider the Markov chain (Xn,n ≥ 0) on state space S = {0,1,2, . . .} with transition probabilities pi j = { (1−α)α j i= 0 and j ≥ 0 (1−α)α j−i+1 i≥ 1 and j ≥ i−1. and all other pi j = 0. Using branching process theory, find the values of α for which (Xn) is transient, null recurrent or positive recurrent. Fully explain your answer. Hint: Relate the Markov chain (Xn) to a “slowed down” branching process, where at any given step a single particle gives birth. Then use the result of Problem 8(b) on Homework 2. 2
欢迎咨询51作业君