ISYE-6501-OAN/O01 -- HOMEWORK 9 – SOLUTIONS
12.1
As a software engineer of a service-oriented company, we try to upsell many products to the customers
and want to display the ads of a product/service within the mobile
...
ISYE-6501-OAN/O01 -- HOMEWORK 9 – SOLUTIONS
12.1
As a software engineer of a service-oriented company, we try to upsell many products to the customers
and want to display the ads of a product/service within the mobile application. Most of the time we
don’t have much idea about the right screen/position of the ad, so it catches most eyeballs and
company can end up selling that product/service. In that case, we use experiment approach such as A-B
testing to place the ad in separate locations and gather the analytics to better understand the most
valuable place to post the ad.
12.2
if (!require("FrF2"))
install.packages('FrF2', repos='http://cran.us.r-project.org')
## Loading required package: FrF2
## Loading required package: DoE.base
## Loading required package: grid
## Loading required package: conf.design
##
## Attaching package: 'DoE.base'
## The following objects are masked from 'package:stats':
##
## aov, lm
## The following object is masked from 'package:graphics':
##
## plot.design
## The following object is masked from 'package:base':
##
## lengths
set.seed(42)
combo <- FrF2(16,10)
combo
This study source was downloaded by 100000834091502 from CourseHero.com on 05-16-2022 06:23:47 GMT -05:00
https://www.coursehero.com/file/69470887/ISYE-6501-hw-9pdf/
Output—
The explanation is fairly simple - each of the factors from A through are binary variables. We have 16
combinations of which factors to keep.
If we look at the first combination, we keep {B, C, G, H, K}.
The combinations were chosen to never repeat between the 16, and also to try to have the most
variability in interactions. For example, both A and B are included only in the 15th and 16th
combinations. Considering that there are 2^10 = 1024 possible combinations, we have some overlap
with other variables.
This study source was downloaded by 100000834091502 from CourseHero.com on 05-16-2022 06:23:47 GMT -05:00
https://www.coursehero.com/file/69470887/ISYE-6501-hw-9pdf/
13.1
a. Binomial
Whether or not a movie will profit - basically, we can think of any binary question for the binomial
distributions that a model such as logistic regression may solve.
b. Geometric
At which day after release that the movie will break even - this may never happen since the movie may
never break even, just like how a bat may never break during the experiment held by the class lecture
example. However, outside of that, we can try to see when a specific production’s movies tend to break
even and find probability parameters based off that.
c. Poisson
The probability of a particular movie ticket (the golden ticket) being bought.
d. Exponential
How many regular tickets are bought before e
[Show More]
.png)
