Quiz 6
Started: Apr 9 at 4:22am
Quiz Instructions
Quiz 6: K353, Spring 2020
Instructions:
• You must submit this assignment by 11:30 pm, Monday, April 13, 2020
• Late submissions receive zero credit.
• There is no time limit with this quiz. You can come back to it later and pick up where you
left off.
• You must submit your own assignment.
• Your assignment is submitted only once you click the "submit" button!
• Do NOT forget to click submit.
• Once you click submit, you will NOT be allowed to submit a different set of answers – you
will only be able to submit once!
• There is no partial credit.
• You are expected to use R to answer all questions.
• If you answered Questions 2-8 but Question 9 (which asks for the R code you used to
answer Question 2-8) is blank, your quiz score will result in, at best, a score of zero.
• If the R code you provided in Question 9 does not match your answers for the rest of the
questions, your quiz score will result in, at best, a score of zero.
5 pts
HTML Editor
Question 1
Accuracy is a numerical measure that is used to evaluate classification
models. Explain how accuracy is calculated. Also, explain why accuracy
may not the best metric to evaluate classification models.
A company that manufactures riding mowers wants to identify the best
sales prospects for an intensive sales campaign. In particular, the
manufacturer is interested in classifying households as prospective owners
or nonowners on the basis of Income (in $1000s) and Lot Size (in 1000 ft2).
The marketing expert looked at a random sample of 24 households, given
in the file RidingMowers.csv (under Datasets).
5 pts
HTML Editor
Question 2
Make sure to include the following lines of code in your R script to load and
modify the dataset:
mower_data <- read.csv('RidingMowers.csv', header = T, as.is = T)
mower_data$OwnershipNum <- mower_data$Ownership
mower_data$OwnershipNum[mower_data$Ownership == "owner"] <-
1
mower_data$OwnershipNum[mower_data$Ownership == "nonowner"]
<- 0
mower_data$OwnershipNum <- as.numeric
(mower_data$OwnershipNum)
In your own words, explain what the above chunk of code does line by line.
Question 3 5 pts
After running the chunk of code in Q2, use all the data (do NOT partition
data) to fit a logistic regression for ownership (OwnershipNum) on two
predictors, Income and Lot size. Based on logistic regression output,
provide the coefficient value for the predictor Lot size (for the equation
where the right-hand side is log of odds). Round your answer to two
decimals.
Question 4 5 pts
Use the logistic regression model in Q3 and cutoff value 0.3 to classify
customers in the dataset as owners and nonowners. Among nonowners,
what is the percentage of households classified correctly? Round your
answer to two decimals.
Question 5 3 pts
Based on the model fitted in Q3, what is the probability that a household
with a $69K income and a lot size of 15,000 ft2 is an owner? Round your
answer to two decimals.
Question 6 2 pts
Assume that it is more important to detect customers who are owners than
detecting customers who are nonowners. Provide a numeric measure for
how good your model in Q3 is at finding the customers who are owners,
assuming that the cutoff is 0.5. Round your answer to two decimals.
Question 7 2 pts
Assume that it is more important to detect customers who are owners than
detecting customers who are nonowners. Provide a numeric measure for
how good your model in Q3 is at finding the customers who are owners,
assuming that the cutoff is 0.3. Round your answer to two decimals.
Question 8 3 pts
Use all the data (do NOT partition data) to fit a logistic regression for
ownership (OwnershipNum) on three predictors, Income, Lot size and
interaction between Income and Lot size. Use this logistic regression model
and cutoff value 0.5 to classify customers in the dataset as owners and
nonowners. What is the accuracy of this model? Round your answer to two