Stat 3126.

Mathematical Statistics 1

Spring 2024

Department of Applied Statistics, Yonsei University

Previous Exam 1

1. [15 points] Consider an experiment that a single die is cast continually until a 6
appears. Let the outcome be a collection of dice faces until a 6 appears.

(a) [5 points] What is the sample space of this experiment?

(b) [5 points] Let Cn denote the event that n rolls are necessary to complete the
experiment. Write down the elements of C3 .
(c) [5 points] Calculate the probability that the die is cast at least twice, i.e.,
∞
!
[
P Cn .
n=2

2. [10 points] A 5-hand poker hand is to be dealt at random without replacement from
an ordinary deck of 52 playing cards, and assume that all 5-hand poker hands are
equally likely.

(a) [5 points] What is the probability of being dealt a diamond flush? (A hand is
said to be a diamond flush if all 5 cards are diamonds.)
(b) [5 points] What is the probability of being dealt two pair? (This occurs when
the cards have denominations, (a, a, b, b, c), where a, b and c are all distinct.)

3. [10 points] Let {Ci } be a decreasing sequence of events, i.e., Ci+1 ⊂ Ci . Define the
c
sets, called rings, as Ri = Ci ∩ Ci+1 for i ≥ 1.

(a) [5 points] Show that P (Ri ) = P (Ci ) − P (Ci+1 ) for i ≥ 1.

(b) [5 points] Using the fact that P (Ac ) = 1 − P (A) for any event A, show that
P (limi→∞ Ci ) = P (∩∞
i=1 Ci ) = limi→∞ P (Ci ).

1
4. [10 points] A high school student Brian received admissions from three colleges: Seoul
National University (SNU), Yonsei University (YU), and Korea University (KU). Con-
sidering all factors that affect his decision, the probabilities of choosing these colleges
are 0.4 for SNU, 0.5 for YU, and 0.1 for KU, respectively. Suppose that the probabil-
ities of finding a girlfriend in the first few months of college are 0.5 given that Brian
goes to SNU, 0.6 given that he goes to YU, and only 0.1 given that he goes to KU,
respectively.

(a) [5 points] What is the probability that Brian finds a girlfriend in the first few
months of college?
(b) [5 points] What is the probability that Brian went to Korea University, given
that he found a girlfriend in the first few months of college?

5. [10 points] The annual premium for a $10000 life insurance policy is $200. That is,
an insurance company charges $200 for a life insurance premium at the beginning of
each policy year, and $10000 is paid to a beneficiary when a person insured under the
life insurance policy dies. Suppose that the insurance policy is effective for the next 3
policy years and the probabilities that a person insured under the life insurance policy
dies are 0.001 in the first year, 0.005 in the second year, and 0.01 in the third year,
respectively.

(a) [5 points] Let X be the insurance company’s net gain (= accumulated premiums
− insured amount) for a single person insured under the life insurance policy.
Find the probability mass function of X.
(b) [5 points] Find the expected net gain of the insurance company for a single person
insured under the life insurance policy. This shows that the insurance companies
never lose as they operate on the fact that they make a profit.

6. [10 points] Let X be a discrete random variable with pmf pX (x) that is positive on
the nonnegative integers and is equal to zero elsewhere.
P∞
(a) [5 points] Show that E(X) = x=0 P (X > x).
(b) [5 points] When the cdf of X is given by FX (x) = 1 −(1/3)x+1 for x = 0, 1, 2, . . . ,
use part (a) to compute the expectation of X.

2
7. [10 points] Consider an ordinary deck of 52 playing cards with 4 suits (black Spades,
red Hearts, red Diamonds, and black Clubs), where each suit consists of 13 cards (Ace,
King, Queen, Jack, 10, 9, 8, 7, 6, 5, 4, 3, and 2).

(a) [5 points] Suppose a deck of 52 cards is shuffled and two cards are randomly
drawn from the deck with replacement. That is, the first card that is randomly
drawn is put back in the deck before randomly drawing the second card from a
shuffled deck of cards. What is the probability that the first card is an Ace and
the second card is a Spade?
(b) [5 points] Suppose a deck of 52 cards is shuffled and two cards are randomly
drawn from the deck without replacement. That is, the first card that is
randomly drawn is not put back in the deck and the second card is randomly
drawn from the remaining 51 cards. What is the probability that the first card is
an Ace and the second card is a Spade?
(Hint: An event that the first card is an Ace is a union of two mutually exclusive
events, a Spade Ace or a non-Spade Ace.)

8. [10 points] Consider a society with three social classes. Each individual may belong
to the lower class (L), the middle class (M ), or the upper class (U ). Suppose that
each individual in the current generation has exactly one child in the next generation.
According to a recent article published in The New York Times, intergenerational
social mobility in the United States is characterized by the following matrix,
Next
U M L
U 0.62 0.38 0.00
Current M 0.36 0.23 0.41
L 0.08 0.50 0.42

That is, a child with a lower-class parent has a 42% chance of remaining in the lower
class, has a 50% chance to rise to the middle class, and has a 8% chance to reach the
upper class. A child with a middle-class parent has a 41% chance of falling to the lower
class, a 23% chance of remaining in the middle class, and a 36% chance of rising to the
upper class. Finally, a child with an upper-class parent has no chance of falling to the
lower class, has a 38% chance of falling to the middle class, and has a 62% chance of
remaining in the upper class. Assume that there are 20% of the current generation in
the lower class, 60% in the middle class, and 20% in the upper class.

3
 (a) [5 points] Compute the probability that the next generation is in the upper class.
(b) [5 points] Compute the probability that the current generation belongs to the
upper class given that the next generation is in the upper class?

9. [10 points] In Korean Nanum Lotto 6/45, one selects 6 numbers between 1 and 45
inclusive, and wins at the lotto if all 6 winning numbers match one’s selected numbers in
any one game. Then the probability of winning at the lotto is calculated as 1/8145060 =
0.0000001228. Suppose that Reina who just turned 20 is going to play one lotto game
every week for the rest of her lifetime, say, 60 years ≈ 3128 weeks. The lotto games
are assumed to be independent every week.

(a) [5 points] For each lotto game, we suppose that Reina either wins at the lotto
or not. To enter each lotto game, Reina has to pay 5 thousand Korean Won
(KRW). When Reina wins at the lotto, however, she can pocket 10 billion KRW
after paying tax. Thus, her net gain for each lotto game is as follows.

Wins at the lotto Does not win at the lotto

9, 999, 995, 000 KRW −5, 000 KRW

Find her expected net gain for each lotto game.

(b) [5 points] Compute the probability that Reina wins at the lotto at least once for
the rest of her lifetime, i.e., 3128 weeks.

10. [20 points] How likely do you think you can find someone who shares the same birthday
as you in a group of n students including you, where n < 365? To find out this, solve
the following questions. Suppose that students are equally likely to be born on any of
the 365 days of the year, ignoring the complication of February 29 birthdays.

(a) [5 points] Find the number of possible ways to have n birthdays with no matches,
so that none of the n students have the same birthday.
(b) [5 points] Compute the probability that at least one pair of n students shares
the same birthday, i.e., the probability of at least one match among n birthdays.
(c) [5 points] Find the number of possible ways that (n − 1) students other than you
do not share birthdays with you, i.e., all of (n − 1) birthdays are different from
your birthday. Assume that your birthday is fixed as January 1st.

4
 (d) [5 points] Compute the probability that at least one student out of (n − 1)
students shares the same birthday as you. Assume that your birthday is fixed as
January 1st.

11. [10 points] At the beginning of college study, 30% were classified as Yonsei University
(YU) students, 20% were classified as Seoul National University (SNU) students, and
50% were classified as Korea University (KU) students. After 4 years, it was determined
that the employment rates of the YU and SNU students were six and three times that
of the KU students, respectively.

(a) [5 points] Let x denote the employment rate of KU students. Use x to express
the probability that a randomly selected college student is employed after 4 years.
(b) [5 points] Find the probability that a randomly selected college student was a
KU student, given that the student is employed after 4 years.

12. [10 points] Let W be a discrete random variable with cdf

(
⌊w⌋
1 − 12 if w > 0
FW (w) =
0 elsewhere,

where ⌊w⌋ is the largest integer less than or equal to w, and let X be a continuous
random variable with cdf
(
1 − e−x − xe−x if x > 0
FX (x) =
0 elsewhere.

(a) [5 points] Determine the pmf of Y = W 2 .

(b) [5 points] Determine the pdf of Z = e−X .

13. [15 points] A mathematical statistics class consists of 10 females and 8 males.

(a) [5 points] What is the probability that a committee of five students, chosen at
random from the class, consists of 3 females and 2 males?
(Note that order is not important.)

Suppose that a committee of five students consisting of 3 females and 2 males has been
selected and the five students stand in a line in random order.

5
 (b) [5 points] What is the probability that the sequence alternates between females
and males, i.e., F M F M F , where F and M stand for female and male, respec-
tively?
(c) [5 points] What is the probability that the three females are not together in the
sequence?

14. [10 points] The use of steroids is banned in professional and organized sports. Con-
sider a urine test for a banned steroid that has a 99% probability of correctly detecting
the use of the steroid when an athlete is actually a steroid user. The test also has
a 90% probability of correctly detecting the non-use of the steroid when an athlete
is actually not a steroid user. It is known that 30% of professional athletes use the
banned steroid. Suppose that the probability that a randomly selected athlete uses
the banned steroid is 30%.

(a) [5 points] Find the probability that the urine test of a randomly selected athlete
results that the athlete uses the banned steroid.
(b) [5 points] Find the probability that a randomly selected athlete is actually a
steroid user given that the urine test results that the athlete uses the banned
steroid.

15. [15 points] Consider the cast of one red die and one white die, and assume that the
two dice can be repeatedly cast under the same conditions. Let X equal the absolute
value of the difference between values showing on the two dice.

(a) [5 points] Find the probability mass function (pmf) of X.

(b) [5 points] Find the cumulative distribution function (cdf) of X.
(c) [5 points] Determine the pmf of Y = (X − 2)2 .

16. [10 points] Suppose there is a 1/3 probability that you drink soju at dinner. Assume
that the decision of drinking soju is made independently each evening.

(a) [5 points] Calculate the probability that you drink soju at 5 consecutive dinners.
(b) [5 points] Calculate the probability that you drink soju at least once out of 5
consecutive dinners.

6
17. [15 points] At the beginning of a study of individuals, 30% were classified as smokers
and 70% were classified as nonsmokers. In the five-year study, it was determined that
the death rate of the smokers was four times that of the nonsmokers, and the survival
rate of the nonsmokers was equal to 0.3 plus that of the smokers.

(a) [5 points] Given that a randomly selected individual was a smoker, calculate the
probability that the participant died over the five-year period.
(b) [5 points] Calculate the probability that a randomly selected individual died over
the five-year period.
(c) [5 points] Given that a randomly selected individual died over the five-year
period, calculate the probability that the participant was a smoker.

18. [15 points] Let X have the probability density function (pdf),

fX (x) = 3(1 − x)2 if 0 < x < 1
0 elsewhere.

(a) [5 points] Find the cumulative distribution function (cdf) of Y = −3 log(1 − X).
(b) [5 points] Find the probability density function (pdf) of Y = −3 log(1 − X).
(c) [5 points] Divide a line segment between 0 and 1 into two parts by selecting
a point x according to the probability density function fX (x). Compute the
probability that the larger segment is at least twice the shorter.

19. [5 points] Let A and B be any two events on a probability space. Express

P ((A ∩ B c ) ∪ (Ac ∩ B))

in terms of P (A), P (B), and P (A ∩ B). (Do not use the Venn diagram.)

20. [15 points] Prostate-specific antigen (PSA) blood test is used to detect prostate cancer
early in men without symptoms. With the PSA test, 90% of men with prostate cancer
will get a positive test result, while 5% of men without prostate cancer will get a
positive test result. Suppose that the prostate cancer incidence rate is 0.001.

(a) [5 points] Calculate the probability of getting a positive test result from the PSA
test.
(b) [5 points] Suppose that independent samples of n men had the PSA test. Cal-
culate the probability that at least one man had a positive test result.

7
 (c) [5 points] Given that a randomly selected man has a positive test result for the
PSA test, calculate the probability that he actually has prostate cancer.

21. [15 points] Let X have the probability density function (pdf),

fX (x) = ce−x/2 if 0 < x < ∞
0 elsewhere.

(a) [5 points] Find the value of c.

(b) [5 points] Find the pdf of Y = eX .
(c) [5 points] Find the cumulative distribution function (cdf) of Y = eX and calcu-
late the median of Y .

22. [5 points] A 5-hand poker hand is to be dealt at random without replacement from an
ordinary deck of 52 playing cards, and assume that all 5-hand poker hands are equally
likely. Compute the probability of being dealt three of a kind. (This occurs when the
cards have denominations, (a, a, a, b, c), where a, b and c are all distinct.)

23. [5 points] Let A and B be any two events on a probability space. Use the law of total
probability to show that

P ((A ∩ B c ) ∪ (Ac ∩ B)) = P (A) + P (B) − 2P (A ∩ B).

24. [10 points] A study of Gi-heung/Hwa-seong residents found that semiconductor work-
ers had significantly increased risk of lung cancer. That is, 20% of those with lung
cancer had worked in a semiconductor plant, while 10% of those without lung can-
cer had worked in a semiconductor plant. Assume that the prevalence of lung cancer
in Gi-heung/Hwa-seong during the study period was 1%. Let S denote the event of
working in a semiconductor plant and let L denote the event of lung cancer.

(a) [5 points] Calculate the probability that a Gi-heung/Hwa-seong resident worked

in a semiconductor plant?
(b) [5 points] Calculate the probability that a Gi-heung/Hwa-seong resident will
have lung cancer, given that he/she worked in a semiconductor plant?

25. [10 points] Let X have the probability density function,

(
2(1 − x) if 0 < x < 1
fX (x) =
0 elsewhere.

8
 (a) [5 points] Find the cumulative distribution function of X.
(b) [5 points] Find the probability density function of Y = 1/X.

26. [10 points] “Better that ten guilty persons escape than that one innocent suffer,”
says English jurist William Blackstone. In criminal law, the ratio 10:1 is known as the
Blackstone’s formulation, i.e., the probability that a guilty person is acquitted should
be (at least) 10 times greater than the probability that an innocent person is convicted.
Suppose the accused is facing a trial in a court, the probability that the accused is
innocent is 20%, and the probability of convicting the accused given that he/she is
innocent is 1%. Following the Blackstone’s formulation, the probability of acquitting
the accused given that he/she is guilty is 10%.

(a) [5 points] Calculate the probability that the accused is convicted.

(b) [5 points] Given that he/she is convicted, calculate the probability that the
accused is innocent.

27. [10 points] Let X have the probability density function,

(
2x if 0 < x < 1
fX (x) =
0 elsewhere.

(a) [5 points] Find the probability density function of Y = 1/X.

(b) [5 points] Find the expectation of X(1 − X).

28. [15 points] Let X be a continuous random variable having the cumulative distribution
function (cdf),

FX (x) = 1 − exp(−3x + a) if x > 2
0 elsewhere.

(a) [5 points] Find the value of a such that the cdf is valid, and find the corresponding
probability density function (pdf) of X.
(b) [5 points] Find the probability density function (pdf) of Y = −3 log(X − 2).

29. [5 points] Suppose that there are 100 light bulbs in a room, 5 of which are defective. An
inspector examines five bulbs, which are selected at random and without replacement.
Find the probability of at least one defective bulb among the five. (Hint: You do not
have to give an exact answer, and you may simplify your answer as much as you can.)

9
30. [5 points] Let X have the probability density function,
(
3x2 if 0 < x < 1
fX (x) =
0 elsewhere.
Compute E[X 3 ].

31. [5 points] Let MX (t) denote the moment generating function of X and let ψX (t) =
′ ′′
log MX (t). Show that ψX (0) = µ and ψX (0) = σ 2 , where µ and σ 2 are the mean and
variance of X, respectively.

32. [5 points] Let X be a discrete random variable with probability mass function (pmf),
(
1/n if x = a1 , . . . , an
pX (x) =
0 otherwise.
Use the Jensen’s inequality to sort the three common means in descending order, where
n
1X
(1) arithmetic mean (AM) = ai
n i=1
n
!1/n
Y
(2) geometric mean (GM) = ai
i=1
n
!−1
1X 1
(3) harmonic mean (HM) = .
n i=1 ai

33. [5 points] Suppose X is a random variable with probability density function

1
fX (x) = e−|x| , − ∞ < x < ∞.
2
Find the moment generating function of X.

34. [5 points] Suppose that both fX (x) and gX (x) are proper probability density functions
for −∞ < x < ∞. The Kullback-Leibler divergence between fX (x) and gX (x) is then
defined to be Z ∞
fX (x)
KL(f ||g) = fX (x) log
dx.
−∞ gX (x)
Use the Jensen’s inequality to show that KL(f ||g) ≥ 0.

35. [5 points] Let X be a positive continuous random variable with moment generating
function denoted by MX (t) for −h < t < h. Show that

P (X ≥ a) ≤ e−at MX (t)

for 0 < t < h and a > 0.

10
36. [10 points] Suppose there are three factories producing the same smartphone: Factory
A, Factory B, and Factory C. Factory A produces 40% of the smartphones, Factory
B produces 20%, and Factory C produces 40%. The defective rates for smartphones
from Factories A and B are three and five times higher than those for Factory C.

(a) [5 points] If a smartphone is randomly selected and found to be defective, what

is the probability that it came from Factory B? Explain why.
(b) [5 points] Suppose that a defective rate for smartphones from Factory C is 0.1.
When a smartphone is randomly selected 10 times with replacement, what is the
probability that at least one defective smartphone is selected? Choose one and
explain why.

i. 1 − 0.2610
ii. 1 − 0.3210
iii. 1 − 0.6810
iv. 1 − 0.7410

37. [10 points] Let X be the total number of casts until we observe a one when we cast
a fair die and have the pmf
 x−1
1 5
fX (x) = , x = 1, 2, . . . .
6 6

(a) [3 points] Compute the expectation of X.

(b) [3 points] Compute the moment generating function (mgf) of X.
(c) [4 points] Find the pmf of Y = X/(X + 1).

38. [10 points] Let X have the probability density function (pdf),

fX (x) = ce−x if 0 < x < ∞
0 elsewhere.

(a) [2 points] Find the value of c.

(b) [4 points] Find the pdf of Y = eX .
(c) [4 points] Find the cumulative distribution function (cdf) of Y = eX and calcu-
late the median of Y .

39. [10 points] Answer each question with True or False.

11
(a) [1 point] (True/False) Suppose you make a bet with a friend where you will
receive $X if you win, but must give $1 to your friend if you lose. You have
decided to join the bet only when the payout to you is $2 (i.e., X = 2). Based on
this threshold, your subjective probability of winning the bet is higher than the
probability of getting a head in a fair coin toss.
(b) [1 point] (True/False) Let C be a set in one-dimensional space and let Q(C) =
R x
C
2 dx. If C = {x : −∞ < x ≤ 0}, then Q(C) = 1/ log 2.
(c) [1 point] (True/False) Let C1 and C2 be subsets of the sample space Ω. If
P (C1 ∪ C2 ) = P (C1 ) + P (C2 ), C1 and C2 are said to be mutually independent.
(d) [1 point] (True/False) Teams A and B are playing a three-game series, where
the team that wins the first two games wins the series. Each game is independent,
and the probability of team A winning each game is 1/3. Given that team A wins
the first game, the probability that team A will win the series is computed as 4/9.
(e) [1 point] (True/False) Let X be a discrete random variable such that pX (x) =
x/10 if x = 1, 2, 3, 4 and 0 elsewhere. Then, P (X ≤ 3.5) = 3/5.
(f) [1 point] (True/False) The mean of a random variable always exists if its vari-
ance is finite.
(g) [1 point] (True/False) Let X be a continuous random variable such that fX (x) =
1 −|x|
2
e for −∞ < x < ∞. Then, the mgf of X is M (t) = (1 − t2 )−1 for |t| < 1.
(h) [1 point] (True/False) Let X be a continuous random variable with P (X <
0) = 0. When E[X] = µ exists, P (X ≥ 3µ) ≤ 1/3.
(i) [1 point] (True/False) Let X be a nondegenerate random variable with mean
µ and a finite second moments. Then µ2 ≥ E(X 2 ).
(j) [1 point] (True/False) Let X be a random variable with probability density
function p(x) and let q(x) be another probability density function with the same
support of X. Then Jensen’s inequality implies that
   Z  
q(X) q(x)
E − log = − log p(x)dx ≤ 0.
p(X) p(x)

12
Solution Keys for Previous Exam 1

1. (a) Ω = {6, (1, 6), (2, 6), (3, 6), (4, 6), (5, 6), (1, 1, 6), . . . }.
(b) C3 = {(1, 1, 6), (1, 2, 6), . . . , (5, 5, 6)}.
(c) Use the rule for complement to calculate the probability.
13 52



2. (a) 5
/
= 0.0004951981.
5

(b) 13

3
4
4
4
52

3 2 2 2 1
/ 5 = 0.04753902.

3. (a) P (Ci ) = P (Ci+1 ∪ Ri ) = P (Ci+1 ) + P (Ri ).

(b)
  ∞
\ 
P lim Ci = P Ci
i→∞
i=1
∞
[ 
= 1−P Cic
i=1
 
= 1−P C1c ∪ R1 ∪ R2 ∪ . . .
i
X
= P (C1 ) − lim P (Rj )
i→∞
j=1
 i n
X o
= lim P (C1 ) − P (Cj ) − P (Cj+1 )
i→∞
j=1
= lim P (Ci )
i→∞

4. (a) Let G denote an event that Brian finds a girlfriend in the first few months of
college

P (G) = P (G|SNU)P (SNU) + P (G|YU)P (YU) + P (G|KU)P (KU)

(b)

P (G|KU)P (KU)
P (KU|G) =
P (G)

5. (a) The possible values X can take on are −9800, −9600, −9400, and 600. Then you
can find the corresponding probabilities.
(b) The expected value of X is 438.6.

13
6. (a)
∞
X
E(X) = xP (X = x)
x=0
= P (X = 1) + {P (X = 2) + P (X = 2)} + {P (X = 3) + P (X = 3) + P (X = 3)} + · · ·
X∞
= P (X > x)
x=0

(b)
∞
X
E(X) = (1 − FX (x)) = 0.5.
x=0

7. (a)

#{Drawing two cards with replacement} = 522

#{First Ace & Second Spade} = 4 × 13

(b)

P (1st Ace & 2nd ♠) = P (1st Ace & 2nd ♠|1st ♠ Ace)P (1st ♠ Ace)
+ P (1st Ace & 2nd ♠|1st non-♠ Ace)P (1st non-♠ Ace)
= P (2nd ♠|1st ♠ Ace)P (1st ♠ Ace)
+ P (2nd ♠|1st non-♠ Ace)P (1st non-♠ Ace)

8. (a)

P (U2 ) = P (U2 |L1 )P (L1 ) + P (U2 |M1 )P (M1 ) + P (U2 |U1 )P (U1 )

(b)
P (U2 |U1 )P (U1 )
P (U1 |U2 ) =
P (U2 )

9. (a)

E(net gain) = −3772.262

(b)

P (winning) = 0.0003839627.

14
10. (a)

365 × 364 × · · · × (365 − n + 1)

(b)

P (at least one match) = 1 − P (no match)

(c)

364n−1

(d)

P (at least one match with your birthday) = 1 − P (no match with your birthday)

11. (a)

P (E) = P (E|Y U )P (Y U ) + P (E|SN U )P (SN U ) + P (E|KU )P (KU )

= 2.9x.

(b)

P (E|KU )P (KU )
P (KU |E) =
P (E)

12. (a)

pW (w) = FW (w) − FW (w − 1)
 w
1
= , w = 1, 2, 3, . . .
2
 y1/2
1
pY (y) = , y = 1, 4, 9, . . . .
2

(b)

dFX (x)
fX (x) =
dx
= xe−x .
fZ (z) = fX (g −1 (z))|J|

15
13. (a) The number of ways of choosing a committee of five students consisting of 3
females and 2 males is
   
10 8
× = 120 × 28 = 3360
3 2
#{3 females and 2 males}
P (3 females and 2 males) =
#{5 students}

(b)
#{F M F M F }
P (F M F M F ) =
#{making a line in random order}

(c)

P (3 females are not together) = 1 − P (3 females are together)

14. (a)

P (+) = P (+|S)P (S) + P (+|S c )P (S c )

(b)
P (+|S)P (S)
P (S|+) =
P (+|S)P (S) + P (+|S c )P (S c )

15. (a) The possible values of the difference between two dice tolls are S = {−5, −4, . . . , 4, 5},
so the possible values of the absolute difference are

S = {0, 1, 2, 3, 4, 5}.

x 0 1 2 3 4 5
P (X = x) 6/36 10/36 8/36 6/36 4/36 2/36

(b)

x 0 1 2 3 4 5
P (X ≤ x) 6/36 16/36 24/36 30/36 34/36 1

(c)

y 0 1 4 9
P (Y = y) 8/36 16/36 10/36 2/36

16
16. (a)

P (A1 ∩ A2 ∩ · · · ∩ A5 ) = P (A1 )P (A2 ) · · · P (A5 )

(b)

P (A1 ∪ A2 ∪ · · · ∪ A5 ) = 1 − P (Ac1 )P (Ac2 ) · · · P (Ac5 )

17. (a) Let π denote the death rate of the nonsmokers.

P (Dc |N S) = P (Dc |S) + 0.3, so we have

π = 0.1.

(b)

P (D) = P (D|S)P (S) + P (D|N S)P (N S)

(c)
P (D|S)P (S)
P (S|D) =
P (D)

18. (a)

P (Y ≤ y) = P (−3 log(1 − X) ≤ y)
= P (X ≤ 1 − e−y/3 )

(b) The support of Y is given by (0, ∞). The inverse of the transformation is X =
1 − e−Y /3 , and the Jacobian of the transformation is J = dx/dy = e−y/3 /3.
dx
fY (y) = fX (x)
dy
(c)
20
P (X < 1/3 or X > 2/3) =
27
19.

P ((A ∩ B c ) ∪ (Ac ∩ B)) = P (A ∩ B c ) + P (Ac ∩ B) because of disjoint events

P (A) = P (A ∩ B) + P (A ∩ B c ) by the law of total probability
P (B) = P (A ∩ B) + P (Ac ∩ B) by the law of total probability

17
20. (a)

A = positive result
B = prostate cancer
P (A) = P (A|B)P (B) + P (A|B c )P (B c )

(b)

Ai = positive result for the ith man

P (at least one positive) = 1 − P (none of them are positive)

(c)

P (A|B)P (B)
P (B|A) =
P (A)

21. (a)
Z ∞ Z ∞
1 = fX (x)dx = ce−x/2 dx
0 0

(b)

0 < X < ∞ ⇒ 1 < Y < ∞

dx
fY (y) = fX (x)
dy

(c)
Z y
1 −3/2
FY (y) = z dz
1 2

22.
13

3
4
4
4

3 1 3 1 1
52

.
5

23.

P ((A ∩ B c ) ∪ (Ac ∩ B)) = P (A ∩ B c ) + P (Ac ∩ B) because of disjoint events

P (A) = P (A ∩ B) + P (A ∩ B c ) by the law of total probability
P (B) = P (A ∩ B) + P (Ac ∩ B) by the law of total probability

18
24. (a)

P (S) = P (S|L)P (L) + P (S|Lc )P (Lc )

= 0.101

(b)

P (S|L)P (L)
P (L|S) =
P (S)
= 0.0198

25. (a)



 0 if x < 0
FX (x) = 2
2x − x if 0 ≤ x < 1

if x ≥ 1.

 1

(b)
(
2(y − 1)/y 3 if 1 < y < ∞
fY (y) =
0 elsewhere.

26. (a)

P (C) = P (C|I)P (I) + P (C|G)P (G)

= 0.722

(b)

P (C|I)P (I)
P (I|C) =
P (C)
= 1/361

27. (a)

X = 1/Y
(
2/y 3 if 1 < y < ∞
fY (y) =
0 elsewhere.

19
 (b)
1
E[X(1 − X)] =
6

28. (a)

a = 6
dFX (x)
fX (x) =
dx
= 3 exp(−3(x − 2)).

(b) The support of Y is given by (−∞, ∞). The inverse of the transformation is X =
2 + e−Y /3 , and the Jacobian of the transformation is J = dx/dy = −(1/3)e−y/3 .
−y/3 −y/3
fY (y) = e−3e , − ∞ < y < ∞.

29. Let C be an event that the at least one light bulb is defective.

P (C) = 1 − P (C c )
95


5
= 1− 100


5
95!95!
= 1−
90!100!

30.
Z 1
3
E[X ] = x3 3x2 dx
0
1
=
2

31.

′ MX′ (t)
ψX (t) =
MX (t)
′′ MX′′ (t)MX (t) − {MX′ (t)}2
ψX (t) =
MX (t)2

20
32.

log(AM) = log(E[X])
log(GM) = E[log(X)]
  
1
− log(GM) = E log
X
 
1
− log(HM) = log E
X
⇒ AM ≥ GM ≥ HM

33.
Z 0 Z ∞
tx 1 x 1
MX (t) = e e dx + etx e−x dx
−∞ 2 0 2
1
=
1 − t2

34.
Z ∞
gX (x)
KL(f ||g) = − fX (x) log dx
−∞ fX (x)
 
gX (x)
= −E log
f (x)
 X 
gX (x)
≥ − log E
fX (x)
= 0

35.
Z ∞
MX (t) = etx fX (x)dx
Z−∞
∞
≥ etx fX (x)dx
a
≥ eta P (X ≥ a)

36. (a)
P (D|B)P (B)
P (B|D) =
P (D|A)P (A) + P (D|B)P (B) + P (D|C)P (C)
5
=
13

21
 (b)

P (D) = 0.26
P (at least one defective) = 1 − P (no defective)
= 1 − (1 − P (D))10

37. (a)
∞  x−1
1 X 5
E[X] = x·
x=1
6 6
= 6

(b)
∞  x−1
X 1
tx 5
MX (t) = e ·
x=1
6 6
et 5et
= , for < 1 or t < log(6/5)
6 − 5et 6

(c)
Y
X =
1−Y
(
2y−1
1 5 1−y
6 6
if y = 21 , 23 , 34 , . . .
pY (y) =
0 elsewhere.

38. (a)

c=1

(b)
dx
fY (y) = fX (x)
dy
(
y −2 if 1 < y < ∞
=
0 elsewhere.

22
 (c)

FY (y) = 1 − y −1
1 − ŷ −1 = 0.5

39. (a) False

(b) True
(c) False
(d) False
(e) True
(f) True
(g) True
(h) True
(i) False
(j) False

23