8 Solved Problems - Assignment 3 | MATH 597, Assignments of Mathematics

Material Type: Assignment; Class: ST-Nonlinear Dynamical Lattices; Subject: Mathematics; University: University of Massachusetts - Amherst; Term: Unknown 1989;

Typology: Assignments

Pre 2010

Uploaded on 08/19/2009

koofers-user-iau
koofers-user-iau 🇺🇸

9 documents

1 / 3

Toggle sidebar

This page cannot be seen from the preview

Don't miss anything!

bg1
Math 597/697: Homework 3
You are encouraged to work in groups. But please write your solutions
yourself. If you have questions about the homework, please ask them in
class, the other students will profit from them too.
1. The Smiths receive the paper every morning and place it on a pile
after reading it. Each afternoon, with probability 1/3 someone takes
all papers in the pile and put them in the recycling bin. Also if ever
there at least five papers in the pile, Mr Smith, with probability one,
take the papers to the bin in the afternoon. Consider the number of
papers in the pile in the evening and describe it with a Markov chain.
(a) What are the state space and the transition matrix?
(b) After a long time what would be the expected number of papers
in the pile?
(c) Assume that the piles starts with 0 papers. What is the expected
time until the pile will have again 0 papers.
(d) Same question as (c) but assume now that the pile starts with 2
papers.
2. Consider an irreducible Markov chain with a finite state space. Let
Mij =E[τj|x0=i] (1)
denote the expected return time to the state j, starting from i.
(a) Analyzing the first step show that
Mij = 1 + X
k
PikMkj (2)
(b) Let πibe the stationary distribution. Multiplying both sides by
πiand summing over ishow that
πj=1
Mjj
.(3)
3. The Ehrenfest urn model, continued from Problem 3 of HWK #2. This
is a Markov chain on {−a,a+ 1, . . . , a 1, a}with nonzero transition
probabilities given by Pii+1 = (ai)/2aand Pii1= (a+i)/2a.
1
pf3

Partial preview of the text

Download 8 Solved Problems - Assignment 3 | MATH 597 and more Assignments Mathematics in PDF only on Docsity!

Math 597/697: Homework 3

You are encouraged to work in groups. But please write your solutions yourself. If you have questions about the homework, please ask them in class, the other students will profit from them too.

  1. The Smiths receive the paper every morning and place it on a pile after reading it. Each afternoon, with probability 1/3 someone takes all papers in the pile and put them in the recycling bin. Also if ever there at least five papers in the pile, Mr Smith, with probability one, take the papers to the bin in the afternoon. Consider the number of papers in the pile in the evening and describe it with a Markov chain.

(a) What are the state space and the transition matrix? (b) After a long time what would be the expected number of papers in the pile? (c) Assume that the piles starts with 0 papers. What is the expected time until the pile will have again 0 papers. (d) Same question as (c) but assume now that the pile starts with 2 papers.

  1. Consider an irreducible Markov chain with a finite state space. Let

Mij = E[τj |x 0 = i] (1)

denote the expected return time to the state j, starting from i.

(a) Analyzing the first step show that

Mij = 1 +

k

PikMkj (2)

(b) Let πi be the stationary distribution. Multiplying both sides by πi and summing over i show that

πj =

Mjj

  1. The Ehrenfest urn model, continued from Problem 3 of HWK #2. This is a Markov chain on {−a, −a + 1,... , a − 1 , a} with nonzero transition probabilities given by Pii+1 = (a − i)/ 2 a and Pii− 1 = (a + i)/ 2 a.

(a) Let μn = E[Xn]. Show that μn = ( a− a 1 )nμ 0 and so limn→∞ μn =

  1. Hint: Use conditional expectation to express μn in terms of μn− 1. (b) Show that

πj =

( 2 a a + j

) 2 −^2 a^ (4)

is the stationary distribution for Xj.

  1. Consider the simple queuing model introduced in class. During each time period exactly a new customer arrives with probability p and no customer arrives with probability 1 − p. During each time period exactly one customer is served with probability q and zero is served with probability (1 − q). The transition probabilites are

P 00 = (1 − p) , P 01 = p Pii− 1 = (1 − p)q Pii = pq + (1 − p)(1 − q) Pii+1 = p(1 − q) .(5)

Determine for which p and q the Markov chain is positive recurrent.

  1. Consider independent trials which result in successs S with probability p and failure F with probability q = 1 − p. We say that a success run of length r happened at trial n if the outcomes in the preceeding r + 1 trials, including the the present trial as the last, were F, S, S, ... S. Now let us denote by Xn the length of the trial run at trial n. This is markov chain with state space { 0 , 1 , 2 , 3 ,.. .}.

(a) Verify that the transition probabilities are pi, 0 = (1 − p) and pii+1 = p (b) Show that 0 is positive recurrent by computing E[τ 0 |x 0 = 0]. (c) To compute the stationary distribution, first use (b) to determine π 0 and then use the equation πP = π to determine π 1 , π 2 ,.. ..

  1. Consider the Markov chain with state space { 0 , 1 , 2 , 3 } and transition probabilities

P =

  

  .^ (6)

Compute