Firstly, as with many hardware-software industries, videogames exhibit network effects in that the value of
purchasing a videogame console as a consumer increases in the number of other consumers who also
decide to purchase that console. Although there i
Sneha Jha
1. Concentrate Industry is certainly an attractive business. Majority of costs incurred in the
concentrate industry are for advertising, promotion, market research and bottler support,
whereas the bottling industry is very capital intensive and
Sneha Jha
Airborne
Airborne had several cost advantage over its rivals. It saved a lot of money by purchasing a vacant airbase
and therefore avoided paying landing fee. It also paid reduced property taxes because of the airports
status as a Community Rein
The main way that Enterprise differs from its rivals is that it does not have offices in the airport.
It instead strategically chooses its locations in a way that 90% of the American population lives
within 15 minutes of its offices. The Enterprise manage
I have answered the exam in 4.and then 1.2.3 fashion
Online personals market experiences extreme competitiveness among the firms participating in it. Firms
like EHarmony, Yahoo Personals and Match rule the roost and offer paid subscriptions to its members
According to the case study, the Ryan brothers intended to run 4 round trips per day with a 44-seat turboprop
with each ticket priced at 98 Irish pounds. For convenience, I am using Irish Pounds as standard metric for
currency. We can calculate the yearl
According to the case study, the Ryan brothers intended to run 4 round trips per day with a 44-seat turboprop
with each ticket priced at 98 Irish pounds. For convenience, I am using Irish Pounds as standard metric for
currency. We can calculate the yearl
In the 1940s, frozen foods were uncommon and little infrastructure was in place to facilitate the sourcing,
production, and distribution of frozen food. In addition, few retailers had freezers in their stores to stock
and sell frozen products. These chal
Betfair is more or less like a traditional betting system with additional opportunities for punters to exploit (described
below). Just like bookmakers need a license to operate, Betfair can operate only where it is permitted to operate.
Bookmakers gauge a
Ownership of a company legally provides the buyer company access to the talent, resources and
competencies of the acquired asset. These can have multiple advantages, ranging from immediate
increases in revenues to improving long term financial outlook to
Pre-MBA Statistics
August 21-25, 2006
Instructor: Arthur G. Korteweg
Email: [email protected]
Introduction
Goals:
Refresh basic math skills
Prepare for Statistics courses (41000 and
41100)
Use of Excel for statistics applications
Notes as reference
Answer Key to Pre-MBA Statistics
Descriptive Statistics
1) Name the correct data type and level of measurement:
a) Quantitative, Discrete, Ordinal
b) Quantitative, Continuous, Ratio (you could think of it as discrete if age is measured in
years)
c) Quanti
Answer Key to Pre-MBA Statistics
Inferential Statistics
1) Go to the class website and open the file romans.xls
a) The sample mean age is X = 41.27 and the sample standard deviation sX = 17.06. The
sample size N = 55.
s
s
The 95% Confidence Interval is X
Answer Key to Pre-MBA Business Statistics
Math Lecture
September 15, 2006
1
Solve the following equations for x
a) x = 1
b) x = 1. Note that 2(x + 1) = 2x + 2 so the question is really the same as (a) after you
work out the brackets.
c) x = 4
d) x = 2 or
Answer Key to Pre-MBA Statistics
Probability Theory
1) Suppose you have the sample space
a) A U B = cfw_1,2,3
b) A B = cfw_2
c) AC = cfw_3 and BC = cfw_1
d) P(A) = P(1) + P(2) = 1/3 + 1/2 = 5/6
P(B) = P(2) + P(3) = 1/2 + 1/6 = 2/3
e) P(A U B) = P(1) + P(
Answer Key to Pre-MBA Statistics
Random Variables
1) Pdf and cdf
a) pdf:
cdf:
X
Prob
X
Cum. Prob
1
1/6
1
1/6
2
1/6
2
1/3
3
1/6
3
1/2
4
1/6
4
2/3
5
1/6
5
5/6
6
1/6
6
1
The mean is 3.5 and the variance is 2.92
b) pdf:
cdf:
X
Prob
X
Cum. Prob
2
1/6
2
1/6
4
1
STATE
AL
AZ
AR
CA
CO
CT
DE
FL
GA
ID
IL
IN
IA
KS
KY
LA
ME
MD + DC
MA
MI
MN
MS
MO
MT
NE
NV
NH
NJ
NM
NY
NC
ND
OH
OK
OR
PA
RI
SC
SD
TN
TX
UT
VT
VA
WA
WV
WI
WY
AREA
51,610
113,909
53,104
158,693
104,246
5,009
2,057
58,560
58,876
83,557
56,399
36,291
56,288
82,
B31104 Pre-MBA Statistics
Class Schedule : August 21-25, 2006, 8:30am-4pm.
Instructor
: Arthur G. Korteweg
Email
: [email protected]
TA
Email
: Tony Tang
: [email protected]
Course Overview:
This preparatory course is intended for those students
Adding Functions Written in C to R
P.Rossi 2/6/05
There may come a time when a function written in R runs so slowly that you must convert
part of the function to some lower-level language such as C or FORTRAN. Typically, you will
profile your code in R an
Notes:
model:
X z 1
Y X w 2
1
~ N (0, )
2
Gibbs sampler:
| .
( , ) | .
| .
In these simple simulations w=1 and z includes an intercept and one variable.
These parameters are never changed:
We have 22 design:
.5 1 ii 1 1 1
.8
4
2
.1
.1
The idea i
Some Useful R Pointers
P. Rossi 1/1/05
Revised 1/11/05
Note: these notes assume a Windows environment. All R commands and objects are the
same under Windows and LINUX but the install procedure and GUI are different.
Obtaining R
Visit http:/cran.r-project.
37904
Week 3 Homework Solutions
1. GHK simulator:
The R-code for the GHK simulator and the program to test it using the C-code are in
the file test_ghk.R in the week 3 directory. One run of each of the versions is
sufficient, provided the seed for the ran
37904
Problem Set 4 - Solutions
Problem 1:
Model
z = X + ,
~ iid N ( 0, 2 I )
y = I ( z > 0)
Priors
2 ~
0 s02
;
2
~ N ( , A1 )
0
The likelihood for this model can be written as
L ( , 2 ) = f ( y | X , , 2 ) = f ( y, z | X , , 2 ) dz = L ( , 2 , z ) dz
37904
Week 2 Homework Solutions
1. Univariate Bayes Regression
a. - see the R code in the file univariate-regression.R in the Week 2 directory. The
values of beta and sigma-squared from the regression are
Parameter
beta0: intercept
beta1: log(PRICE) coeff
rtrun=
function(mu,sigma,a,b)cfw_
# function to draw from univariate truncated norm
# a is vector of lower bounds for truncation
# b is vector of upper bounds for truncation
#
FA=pnorm(a-mu)/sigma)
FB=pnorm(b-mu)/sigma)
mu+sigma*qnorm(runif(length(mu)*(FB
37904
Week 1 Homework
1. A list structure is a very good way to store panel data. Each cross-sectional unit can
be a different element in the list. Panel data can be simply a list of lists. This allows
for a different number of observations per panel unit
ldatadf=list(A=df[df[,1]="A",2:4],B=df[df[,1]="B",2:4])
lmout=NULL
for (i in 1:2) cfw_
lmout[i]=lm(Y ~ X1+X2,data=ldatadf[i])
print(lmout[i])
y=as.numeric(dfmat[,2])
X=matrix(as.numeric(dfmat[,3:4]),ncol=2)
X=cbind(rep(1,nrow(X),X)
XpXinv=chol2inv(cho
37904
Week 3 Homework
As always, show both your code and your results. Learning to work with simulation output
and display it in a useful way is important too!
1. Code up your own version of the GHK method using R. Check your method using
the c version of