6.041/6.

431 Spring 2006 Final Exam

Monday, May 22, 1:30-4:30PM

DO NOT TURN THIS PAGE OVER UNTIL

YOU ARE TOLD TO DO SO

Name: Question Score Out of

0
1 42
Recitation Instructor:
2 30
3 28
TA: Your Grade 100

• You have 3 hours to complete the exam.

• At the end of the exam period you will turn in this exam packet, and 2 blue books.
Question 1 will be answered in the exam packet, while question 2 and 3 will be answered
in their own respective blue books.
• Question 1 is multiple choice, no partial credit is given.
• For questions 2 and 3, you should concisely indicate your reasoning and show all relevant
work. Grades will be based on our judgment of your level of understanding as reflected by
what you have written. Writing the correct answer without clearly showing your under­
standing of how to arrive at the answer will receive no credit.
• Answers in the form of an arithmetic expression (sums, products, ratios, factorials) should
be fully reduced to receive full credit.
• This is a closed-book exam except for three double-sided 8.5” by 11” formula sheets.
• Calculators are allowed.
• Be neat! If we can’t read it, we can’t grade it.
 Massachusetts Institute of Technology
Department of Electrical Engineering & Computer Science
6.041/6.431: Probabilistic Systems Analysis
(Spring 2006)

Problem 0:
Write your name, your recitation instructor’s name, and TA’s name on the cover of the exam booklet

and your 2 blue books. Include the question number on the cover of the blue book.

Problem 1: (42 points)

Multiple Choice Questions: There is only one correct answer for each question listed below, please

clearly indicate the correct answer. There will be no partial credit given for multiple choice questions,

thus any explanations will not be graded. Each question is equally weighted at 3.5 points each.

(I) Alice and Bob each choose a number independently and uniformly at random from the interval
[0, 2]. Consider the following events:
A : The absolute difference between the two numbers is greater than 1/4.
B : Alice’s number is greater than 1/4.
Find P(A ∩ B).
49
(a)
128

75

(b)
128

85

(c)
32

85

(d)
128

(II) There are m red balls and n white balls in an urn. We draw two balls simultaneously and at
random. What is the probability that the balls are of different color?
mn

(a)
(m + n)(m + n − 1)

2mn
(b)
(m + n)(m + n − 1)

(c)
mn
2
(d)
mn

Page 2 of 8
 Massachusetts Institute of Technology
Department of Electrical Engineering & Computer Science
6.041/6.431: Probabilistic Systems Analysis
(Spring 2006)

(III) 20 black pebbles are arranged in 4 rows of 5 pebbles each. We choose 4 of these pebbles at
random and color them red. What is the probability that all the red pebbles lie in different
rows?
54 · 5! · 15!

(a)
20!

5
4 · 5! · 15!

(b)
20!

4
5 · 4! · 16!

(c)
20!

45 · 4! · 16!

(d)
20!

(IV) We have two light bulbs, A and B. Bulb A has an exponentially distributed lifetime with mean
lifetime 4 days. Bulb B has an exponentially distributed lifetime with mean lifetime 6 days. We
select one of the two bulbs at random; each bulb is equally likely to be chosen. Given that the
bulb we selected is still working after 12 hours, what is the probability that we selected bulb A?
1
(a) 1

1 + e 24

1
(b)
1 + e
−1
1

(c)
1 + e
1
1

(d) 1

1 + e− 24

(V) A test for some rare disease is assumed to be correct 95% of the time: if a person has the
disease, the test results are positive with probability 0.95, and if the person does not have the
disease, the test results are negative with probability 0.95. A random person drawn from a
certain population has probability 0.001 of having the disease. Given that the person just tested
positive, what is the probability of having the disease?

(a) 0.00095
(b) 0.001
(c) 0.0187
(d) 0.128

Page 3 of 8
 Massachusetts Institute of Technology
Department of Electrical Engineering & Computer Science
6.041/6.431: Probabilistic Systems Analysis
(Spring 2006)

(VI) Let X be a random variable with

2e−2x if x ≥ 0
fX (x) = .
0 otherwise

Let Y = 2X − 4. Find fY (0).

(a) 1 − e−4
(b) 2e−4
(c) 2e−8
(d) e−4

(VII) X and Y are independent random variables, with



 if y = 1
3e−3x if x ≥ 0  1/2
fX (x) = , pY (y) = 1/2 if y = 0 .
0 otherwise 
 0 otherwise

Let Z = X + Y . What is the transform of Z, MZ (s), evaluated at s = 1?

(a) 3(1 + e)/2

(b) 3(1 + e)/4
(c) 3e(1 + e)/2
(d) 3e(1 + e)/4

(VIII) A number p is drawn from the interval [0, 1] according to the uniform distribution, and then a
sequence of independent Bernoulli trials is performed, each with success probability p. What is
the variance of the number of successes in k trials? Note k is a deterministic number.
k
(a)
2

k(1 + k)

(b)
4

k(2 + k)

(c)
12
k
(d)
4

Page 4 of 8
 Massachusetts Institute of Technology
Department of Electrical Engineering & Computer Science
6.041/6.431: Probabilistic Systems Analysis
(Spring 2006)

(IX) A police radar always over-estimates the speed of incoming cars by an amount that is uniformly
distributed between 0 and 5 mph. Assume that car speeds are uniformly distributed between
60 and 75 mph and are independent of the radar over-estimate. If the radar measures a speed
of 76 mph, what is the least squares estimate of the actual car speed?

(a) 72.5 mph

(b) 73.0 mph
(c) 73.5 mph
(d) 72.9 mph

(X) You are visiting the Killian rainforest, when your insect repellant runs out. This forest is
infested with lethal mosquitoes, whose second bite will kill any human instantaneously. Assume
mosquito’s land on your back and deliver a vicious bite according to a Bernoulli Process. Also
assume the expected time till the first bite is 10 seconds. If you arrive at the forest at t = 0,
what is the probability that you die at exactly t = 10 seconds?
99
(a)
1010
98
(b) 10
10
   8
9 1 9
(c)
1 10 10
1
(d)
2

(XI) In order to estimate p, the fraction of people who will vote for George Bush in the next election,
you conduct a poll of n people drawn randomly and independently from the population. Your
estimator Mn is obtained by dividing Sn , the number of people who vote for Bush in your
sample, by n, i.e., Mn = Sn /n. Find the smallest value of n, the number of people you must
poll, for which the Chebyshev inequality yields a guarantee that

P(|Mn − p| ≥ 0.01) ≤ 0.01 .

(Use the fact that p(1 − p) ≤ 1/4 for any p ∈ [0, 1].)

(a) 500,000
(b) 250,000
(c) 25,000
(d) 50,000

Page 5 of 8
 Massachusetts Institute of Technology
Department of Electrical Engineering & Computer Science
6.041/6.431: Probabilistic Systems Analysis
(Spring 2006)

(XII) On each day of the year, it rains with probability 0.1, independent of every other day. Use the
Central Limit Theorem (CLT) (with no half step) to approximate the probability that, out of
the 365 days in the year, it will rain on at least 100 days. In the answers below, Φ denotes the
standard normal CDF.

640
(a) 1 − Φ √
3 365

640
(b) Φ √
3 365

635
(c) Φ √
3 365

635
(d) 1 − Φ √
3 365

Page 6 of 8
 Massachusetts Institute of Technology
Department of Electrical Engineering & Computer Science
6.041/6.431: Probabilistic Systems Analysis
(Spring 2006)

Problem 2: (30 points)

Please write all work for Problem 2 in your first blue book. No work recorded below will be graded.

Each question is equally weighted at 6 points each.

You take a safari trip to the Porobilati game reserve. A highlight of the game reserve is the Poseni
river where one can watch deer and elephants coming to drink water. Deer come to the river according
to a Poisson process with arrival rate λd = 8 per hour; elephants come to the river according to an
independent Poisson process with arrival rate λe = 2 per hour. On the first day of your safari, you
reach the Poseni river early in the morning hoping to see some elephants. Assume that deer and
elephants are the only kinds of animals that visit this river.

(a) Let N be the number of animals you see during the first 3 hours. Find E[N ] and var(N ).

(b) What’s the probability of seeing your 3rd elephant before your 9th deer?

(c) At the end of 3 hours, you have seen 24 deer but no elephants. What is the probability that you
will see an elephant in the next hour?

(d) It is now 6 hours since you started watching. You have seen 53 deer but there is still no sight of
an elephant. How many more deer do you expect to see before you see your first elephant?

(e) Unfortunately, you forgot your camera in your lodge the first day. You come back with your
camera the next day and plan to stay at the river until you’ve clicked a picture of both a deer and
an elephant. Assume you’re always able to click a picture as soon as an animal arrives. How long
do you expect to stay?

(f) Your friend, a wildlife photographer, is fascinated by the stories of your trip, and decides to take
the safari herself. All excited, she arrives at the Poseni river, and is prepared to stay as long as
required in order to get lots of beautiful pictures. Unfortunately, she dropped her camera in a bush
on the way, as a result of which the camera is not functioning properly; each time she clicks, there
is a probability 0.1 that the camera will produce a faulty picture, independently of everything
else. Assume she clicks one picture each time an elephant or deer arrives at the river, as soon as
it arrives. Let X be the time (in hours) until she gets her third successful picture of an elephant.
Find the PDF of X.

Page 7 of 8

 Massachusetts Institute of Technology

Department of Electrical Engineering & Computer Science
6.041/6.431: Probabilistic Systems Analysis
(Spring 2006)

Problem 3: (28 points)

Please write all work for Problem 3 in your second blue book. No work recorded below will be

graded. Each question is equally weighted at 4 points each.

The MIT football team’s performance in any given game is correlated to its morale. If the team has
won the past two games, then it has a 0.7 probability of winning the next game. If it lost the last
game but won the one before that, it has a 0.4 probability of winning. If it won its last game but lost
the one before that it has a 0.6 probability of winning. Finally if it lost the last two games it has only
a 0.3 probability of winning the next game. No game can end up in a draw. Consider a starting time
when the team has won its preceding two games.

(a) Graph the minimum state Markov chain model for the MIT football team’s performance. Be sure
to label all transition probabilities, and clearly indicate what state of the world each Markov state
represents. You may find it convenient for the below questions to label each state with an integer.

(b) Find the probability that the first future loss will be followed by another loss.

(c) Let X be the number of games played up to, but not including the first loss. Find the PMF of X.

(d) Either evaluate the steady-state probabilities or explain why they do not exist.

(e) Find a good approximation to the probability that the team will win its 1000th game, given that
the outcomes of games 1000 and 1001 are the same. Clearly state any assumptions you use in the
approximation.

(f) Let T be the number of games up to and including the team’s 2nd consecutive loss (i.e. 2 losses
in a row). Write a system of equations that can be used to find E[T ].

(g) The coach decides that he will withdraw the team from the competition immediately after the
team’s 3rd consecutive loss (i.e. three losses in a row). Let N be the number of games that the
team will play in the competition? Write a system of equations that can be used to find E[N ].

Page 8 of 8