-
MA/STAT 519: Introduction to Probability
Fall 2018, Mid-Term Examination
Instructor: Yip
• This test booklet has FOUR QUESTIONS, totaling 80 points for
the whole test. You have
75 minutes to do this test. Plan your time well. Read the
questions carefully.
• This test is closed book, with no electronic device. One
two-sided-8× 11 formula sheet
is allowed.
• In order to get full credits, you need to give correct and
simplified answers and explain in
a comprehensible way how you arrive at them.
Name: (Department: )
Question Score
1.(20 pts)
2.(20 pts)
3.(20 pts)
4.(20 pts)
Total (80 pts)
1
Answerkey
-
1. Let X and Y be two independent discrete random variables. For
each of the following cases,
compute the conditional distribution of X given X + Y , i.e.
find
P(
X = i∣
∣
∣X + Y = j
)
If possible, relate the conditional distribution to some common,
i.e. well known, distribution.
(a) X is Poisson with parameter λ and Y is Poisson with
parameter µ.
(b) X is Binomial with parameter n and p and Y is Binomial with
parameter m and p.
(c) X and Y are Geometric with parameter p.
2
PCX.it tYjJ P X i XtYj1PXtY jJPlX i Y ji7PlXtYjp X i PH jP XtY
j
a Xv Poisson Y PoissonfmlYn PoissonHtml
t xii i s t.is ic si
i
BinCj.fII
-
You can use this blank page.
3
Bin Imp r.Binlm.plismIals.Y
sb 2HY Bin fntm.pl
Tiffany Hypergeometric
Cc Xngeomcp Yu geomlp
HY Neg Bink p
814814
i of Pfuniform doesnotdependoni
-
2. Suppose n balls are distributed at random into r boxes in
such a way that each ball chooses
a box independently of each other. Let S be the number of empty
boxes. Compute ES and
V ar(S).
(Hint: Consider the random variables Xi (for i = 1, 2, . . . ,
r) which equals 1 if the i-th box is
empty and 0 otherwise. Related S and the Xi’s.)
5
f Xt Xu Xr
ES EX Ext E Xr
EX 2x PIX D to xp o
PALED 14154 no A ballsBox List no ofboxes total no ofempty
tochoose from boxes
Hence Ef r n
EH E E XiE Xi t jXiXjEl Xi gEHiXj
-
You can use this blank page.
6
Since Xi o Y Xi Xinoofboxes toEND E Xi choosefrom
Exit P Xi hXj D no ofmy ballstotal no ofpox i I are
emptyboxes
Hence E rfrfjntrlr.it r fnVouCSJ
EC5 fESJ
rfr rYi rcr DfrIY EIHT
-
3. McDonald’s newest promotion is putting a toy inside every one
of its hamburger. Suppose
there are N distinct types of toys and each of them is equally
likely to be put inside any of
the hamburger. What is the expected value and variance of the
number of hamburgers you
need to order (or eat) before you have a complete set of the N
toys.
(Hint: consider the number of hamburgers you need to order (or
eat) in between getting one
and two dinstinct types of toys, two and three distinct types of
toys, and so forth.)
8
yrTF1 T2 Tz
T burgers
71 1
TegeomfftTs geomfEN
tag az
EIe.iYxLetS
Totalno.ofburgers purchasedThen 5 7 31 1 TN
-
You can use this blank page.
9
EfgeomIp ValGeomlpD opt
ES ET t EE t t ETNIt tf Yzx NyN fix It in
VanB Vat Var th t Vatu9since the Ti's are independent
o i t FIT
ftp.i f piHI
N Tiff
-
4. Consider a box with M white balls and N black balls. You are
asked to get n balls (at
random) out of the box without replacement. Let X be the number
of white balls obtained.
(a) Find the probability distribution of X.
(b) Find the expectation and variance of X.
(Hint: imagine that you get the balls sequentially, one by one.
Introduce the indicator
functions Xi = 1 for i = 1, . . . n defined as Xi equals one if
i-th ball is white and zero
otherwise. Relate X and the Xi’s.)
(c) Now suppose M and N tends to infinity such that MM+N
−→ p. Derive and identify the
limiting probability distribution of X.
(d) Under the same limiting procedure, find the limiting
expectation and variance of X.
11
sN k
Iget n balls
Piti Miyazaki Hypergeometric
b X At Kt Xn
EX EX Ext tEXn
-
You can use this blank page.
12
E Xk PIX memoofwhiteballs
MtNr total no ofballsHence the ith djthEX MEX nM_ balls are
white
Mtn
E I EXT t.EE iXjPlXi D IZ.gPlXi hXj D
mFntiEicm'InfinityTuffy mln DMCMMTN MTN t
vaixt ECXD CEXJ
nmfntntym7MY g taking
-
You can use this blank page.
13
cos Plait 7 Fi1
MIM 1 M ite NIN DIN D IN ostiaT.cn
MtN fMtN nt
I
jMCM D fatia n'terms
NIN D IN n in e terms
jMtN MtNntinterns
H hht cnn.inI fmMiniti x
H H InfinitiM N xw
f p 7 piqn imm arspB.ir n p
mNTN si p f
-
Note The answer also makes intuitive senseas M N to it does not
makeany difference whether it is with orwithout replacement
s d Hence
EX np Vai k npofThe above can also becheckeddirectlyfrom the
formula
EX nM_Man np
vaixt.mn o nfmInYimmnI CmFnT
n p t n in 1 p2 ripnp np npftp.J npq