Professional Documents
Culture Documents
SS-1. The sample space for a random experiment is shown as the shaded area in the figure below.
All outcomes in S are equally likely. A random variable Y is set equal to the y-coordinate of
an outcome s = (x, y).
(a) Find the distribution function of Y .
(b) Determine the type (discrete, continuous, mixed) of the random variable Y .
y
1
1 x
SS-2. A random variable X has the distribution function shown in the figure:
1
FX (x)
0.5
0.25
x 4
1
0,
x + 1 (x 1),
fX (x) = 21 4
4,
0,
x<0
0x1
1<x3
x>3
(c) Give an equation for the probability P(0.5 < X 2) in terms of the distribution function
for X. [ FX (2) FX (0.5)]
(d) Give an equation for the probability P(0.5 < X 2) in terms of the density function
for X.
Answer:
Z 2
0.5
fX (x)dx
(e) Give the numerical value of the probability P(1 < X 2). [1/4]
(f) Give the numerical value of the probability P(1 X 2). [1/2]
SS-3. Problem 2.2 from Stark and Woods:
In a restaurant known for its unusual service, the time X, in minutes, that a customer has
to wait before he captures the attention of a waiter is specified by the following distribution
function:
2
x
2 ,
4,
FX (x) = 21 ,
,
20
1,
0x1
1x2
2 x 10
10 x 20
x 20
fX (x)
c
|x|
, |x| a
a2
0,
otherwise
1
a
FX (x) =
0,
(x+a)2 ,
a x < 0
1,
0x<a
xa
2a2
2
1 (ax)
2 ,
2a
x < a
(c) Find b such that P [|X| < b] = 1/2. [Ans:] b = 1 22 a
SS-5. Let fX (x) = A(x + 1)k u(x), where k > 1 is some parameter of the density.
(a) Determine the value of A (in terms of k) that makes fX (x) a valid density function.
[A = k 1]
(b) Determine the distribution function for X.
Answer:
(
1 (x + 1)k+1 , x 0
FX (x) =
0,
otherwise
1. A student is randomly selected from EEL 5544. The student has previously taken a probability
course with probability 0.35, has previously taken only a statistics course with probability
0.4, and has taken neither a probability nor a statistics course with probability 0.25.
If a student has previously taken a probability course, then he has a 0.55 probability of
getting an A in EEL 5544 and a 0.45 probability of getting a B.
If a student has previously taken only a statistics course, then he has a 0.3 probability
of getting an A in EEL 5544 and a 0.7 probability of getting a B.
If a student has previously taken neither a probability nor a statistics course, then he
has a 0.8 probability of getting an A in EEL 5544 and a 0.2 probability of getting a B.
(a) Find the probability that a randomly selected student from EEL 5544 gets an A in the
course.
(b) Find the probability that a randomly selected student from EEL 5544 gets a B in the
course.
(c) Given that a student got an A in EEL 5544, what is the probability that the student had
previously had a probability class?
(d) Given that a student got an B in EEL 5544, what is the probability that the student had
previously had a statistics class?
(e) Find the MAP decision rules for which class a student is most likely to have taken (i.e.,
probability, only statistics, or neither) given the letter grade in EEL 5544.
2. Consider the function
(
cx2 , 1 x 1
fX (x) =
0,
otherwise
For a discrete random variable, it is the probability mass function, treated as a function
of the parameter.
For a continuous random variable, it is the probability density function, treated as a
function of the parameter.
For example, suppose that conduct N Bernoulli trials and observe k successes. What is
the ML estimate of the probability of success p? Let X be the resulting Binomial random
variable with parameters N and p with probability mass function given by
(
n x
nx , x = 0, 1, . . . , n
x p (1 p)
PX (x; p) =
0,
otherwise
Here, I am showing p in the arguments of PX (x; p) because it is an unknown that we wish to
estimate from an observation of X.
Then given k successes, the ML estimate of p is the value of p that maximizes PX (k; p). Let
p be the ML estimate of p. Then we can find p by differentiating PX (k; p) with respect to p
and setting the result equal to zero.
n k
d
nk
p (1 p)
= 0
dp k
kpk1 (1 p)nk pk (n k)(1 p)nk1 = 0
k(1 p) p(n k) = 0
p = k/n
So, p = k/n.
Using this technique, find the following ML estimates:
(a) The number of arrivals at a telephone switch during a one minute period is modeled
as Poisson with parameter . You monitor a one minute period and observe 20 calls.
What is the ML estimate for ? [Ans: 20]
(b) If it takes 10 transmissions to send a packet through a noisy channel in a system that
uses ARQ, what is the ML estimate for the probability of packet success? [Ans: p=
0.1]
(c) The received amplitude of an RF signal is modeled as a Rayleigh random variable with
parameter 2 . If the observed amplitude is 0.5, what is the ML estimate for 2 ? [Ans
2 = 0.125]
3. Problem 3.90 from Leon-Garcia:
An inspector selects every nth item in a production line for a detailed inspection. Suppose
that the time between item arrivals is an exponential random variable with mean 1 minute,
and suppose that it takes 2 minutes to inspect an item. Find the smallest value of n that with
a probability of 90% or more, the inspection is completed before the arrival of the next item
that requires inspection.
(Note that if the time between arrival is an exponential random variable with mean 1 minute,
then the number of arrivals in a time period T minutes is a Poisson random variable with
parameter (mean) T .)
4. X is a random variable with density given by
1 < x < 0
cx ,
2
fX (x) = c(1 x ), 0 x < 1
0,
otherwise,
where c is a constant to be determined.
(a) Find the value of c to make fX (x) a valid density function.
(b) Is X a continuous, discrete, or mixed-type random variable?
(c) Find the distribution function of X. Be sure to specify its value for all real x.
(d) Find P(X 2 > 1/4).
5. A Gaussian random variable Y has mean zero and variance 2 . Find 2 and k such that
P(Y k) = P(Y > k) = 0.05 and the following three events have equal probability:
k < Y 1
1 < Y 1
1<Y k
6. A professor makes an exam for which he anticipates the grades will be Gaussian random
variables with mean 80 and variance 100.
(a) Suppose the professor want the probability of a student getting an A, B, or C to be
given by
A: 0.3
B: 0.6
C: 0.1
Let X denote an exam grade. Assume the grading scheme is of the standard form:
A: X A
B: B X < A
C: X < B
Find the thresholds A and B .
(b) If the true mean and variance of the grades are 82 and 81, respectively, find the probability
that a randomly selected student got an A on the exam (based on the grading scale
developed in part (a)).
(c) If a student is confident that he made at least an 80 on the exam, what is the probability
that the student made an A on the exam? (Use the grading scale from part (a), but the
mean and variance given in part (b).)
(d) Suppose a student feels confident that his exam score is at least . For what value of
should the student be 90% confident that she got an A? (Again, use the grading scale
from part (a), but the mean and variance given in part (b).)