

Study with the several resources on Docsity
Earn points by helping other students or get them with a premium plan
Prepare for your exams
Study with the several resources on Docsity
Earn points to download
Earn points by helping other students or get them with a premium plan
Material Type: Assignment; Class: Probability with Engrg Applic; Subject: Electrical and Computer Engr; University: University of Illinois - Urbana-Champaign; Term: Spring 2002;
Typology: Assignments
1 / 3
This page cannot be seen from the preview
Don't miss anything!


1. [“…From the town of Bedrock, They’re a page right out of history…”] Fred suggests that he and Wilma play a game in which they will take turns tossing a fair coin; the first one to toss a Head wins. Fred proposes that he will toss first. Wilma agrees to this, but, having taken ECE 313, she knows that she is at an disadvantage. So, she demands that in succeeding games, the loser of the previous game gets to toss first. For n ≥ 1, let pn and qn = 1 – pn respectively denote the probabilities that Fred and Wilma win the n-th game. We saw in class that p 1 = 2/3 > q 1 = 1/3,
(a) Use the theorem of total probability to show that p 2 = 4/9 < q 2 = 5/9. More generally, show that for n ≥
2, pn = (2/3) – (1/3)pn-1 and qn = (2/3) – (1/3)qn-1 and use these difference equations to find pn and qn in terms of p 1 and q 1. (If you never learned in Math 285/286 or Math 315 (or ECE 310) how to solve difference equations, assume that for all values of n, pn can be expressed as a + bαn. Substitute into the above difference equation and solve for α and a; the value of b is obtained from the “initial condition” p 1 = 2/3. Repeat for qn — it is the same difference equation but the “initial condition” q 1 = 1/3 is different.)
(b) What is the limit as n → ∞ of pn and qn? Is this game asymptotically fair?
(c) Fred now proposes that instead of the first one to toss a head winning the game, the first one who matches the previous toss (whether the previous toss is part of the current game or the last toss of the previous game) wins. Wilma accepts but generously insists that, as before, Fred still toss first (so that the poor schmuck has no previous toss to match on his first toss!). What are p 1 and q 1 now? p 2 and q 2? p 3 and q 3? Is this game asymptotically fair? Assume as before that the loser of each game tosses first in the next game.
2. [“Reach out and touch someone…”] The probability that you can hear the sound of a pin dropping onto a table during a long-distance telephone call from Urbana to Hollywood is p 0 if the call is being carried over the AT&T network and p 1 if the call is being carried over the Sprint network. Assume that both p 0 and p 1 are quite small and that p 1 > p 0. You call Miss Candice Bergen from a payphone owned by Sleazo Telecom Corporation which happens to lease its long-distance lines either from AT&T or Sprint (but you don’t know which!), and she very graciously agrees to drop pins one by one onto a table until you hear the sound of one dropping. Thus, if the first pin that you hear is the X th, then X is a geometric random
variable with parameter p 0 or p 1 accordingly as the call is being carried by AT&T or Sprint. Suppose you hear Miss Bergen counting “One, two, three, … ” as she drops the pins, but you don’t hear the sound made by the pins as they land on the table until she says “thirtyfour” and you finally hear the sound of the pin dropping onto the table. (a) Let H 0 and H 1 denote the hypotheses that the call is being carried by AT&T and Sprint respectively. What
is the likelihood ratio LR for the observation that X = 34? (b) The maximum-likelihood decision compares LR to the threshold 1 and announces in favor of H 0 and H 1
according as LR < 1 or LR > 1. Show that this decision rule can be expressed in terms of a threshold test on the observed value of X. (c) If p 1 = 0.04 and p 0 = 0.02, what is the maximum-likelihood decision when X = 34?
(d) AT&T is the lessor of 95% of all long-distance telephone lines while Sprint is the lessor of the remaining 5%, and thus it is reasonable to assume that P(H 0 ) = 0.95. What is the Bayesian (minimum-error- probability) decision when X = 34? (e) Noncredit exercise: Whom do you think is my long-distance carrier?
3. [“I’m leaving on a prop plane”] Consider again Problem #3 of Problem Set #5 in which 15 of the 105 passengers who hold reservations are arriving in Chicago on a connecting flight. If the connecting flight is on time, all 15 show up for the flight to Champaign; else, none of the 15 shows up. Let Y denote the number of nonconnecting passengers who actually show up for the flight. Let H 0 denote the hypothesis that the connecting flight is late, and H 1 the hypothesis that the connecting flight is on time. It is reasonable to assume that the pmf of Y is the same regardless of which hypothesis is true, and hence we model Y as a binomial random variable with parameters (90, 0.9). On the other hand, X , the total number of passengers showing up for the flight, equals Y if H 0 is true, while if H 1 is true, then X = 15 + Y , and thus the pmf of X does depend on which hypothesis is true. (a) Suppose that the gate agent observes that X = 86. What is P{ X = 86} when H 0 is the true hypothesis?
What is P{ X = 86} when H 1 is the true hypothesis? What is the value of the likelihood ratio when X = 86, and what is the agent’s maximum-likelihood decision as to whether the connecting flight is late? (b) Repeat part (a) for the case when the gate agent observes that X = 96. (c) The gate agent knows that π 0 = P{H 0 is the true hypothesis} = 2/3. For each of the two observations
considered in parts (a) and (b), what is the agent’s MAP (or Bayesian or minimum-probability-of-error} decision as to whether the connecting flight is late? (d) What is the probability that all passengers who show up get a seat? Given that all passengers who showed up got a seat, find the (conditional) probability that the connecting flight was late.
4. [“It a’in’t about bipartisan politics; it’s about …”] The Senate of a certain country has 100 members consisting of 43 Conservative Republicans, 21 Conservative Democrats, 12 Liberal Republicans, and 24 Liberal Democrats. Before each vote, the groups caucus separately. Each group decides independently of the other groups whether to support or oppose the motion. All members of the group then vote in accordance with the caucus decision. For those who think that this is the way politics works, I have this beautiful skyscraper on Wacker Drive in Chicago that I am willing to sell to you at a bargain price… (a) Let A, B, C, and D respectively denote the events that the four groups vote for a spending plan that will lead to 50% increase in the DoD budget over the next two years. Suppose that the probabilities of these independent events are P(A) = 0.9, P(B) = 0.6 P(C) = 0.5 and P(D) = 0.2. What is the probability that the bill passes? (b) The President vetoes the bill as a budget-breaker. Let E, F, G, and H respectively denote the independent events that the four groups support the motion to override the veto. If these events have probabilities P(E) = 0.99, P(F) = 0.4, P(G) = 0.6, and P(H) = 0.1, what is the probability that the motion to override the veto passes? Political innocents are reminded that a simple majority (51 or more votes) is required to pass a bill, and a two-thirds majority (67 or more votes) to override a veto.
5. [“Tennis, anyone?”] Consider the following simplified model for a game of tennis. On each serve, let p denote the probability that player A wins the point, and q = 1–p the probability that player B wins the point. Assume that the outcome of each serve is independent of all others. Player A wins the game if the score reaches 4–0, 4–1, or 4–2, while B wins the game if the score reaches 2–4, 1–4, or 0–4. Else, the score reaches 3–3 (called deuce) and from this point onwards, the game continues until one player is two