Sample Test II
Hello, dear friend, you can consult us at any time if you have any questions, add WeChat: daixieit
Sample Test II
Question 1. (2 + 4 + 2 = 8 marks)
Bottega Industries is considering purchasing an insurance policy for its new office building in Chicago, USA. The policy has an annual cost of $10,000. If Bottega Industries does not purchase the insurance and minor fire damage occurs, a cost of $100,000 is anticipated; the cost if major or total destruction occurs is $200,000. The costs, including the state-of-nature probabilities, are as follows:
|
Damage |
||||||
|
Decision Alternative None,s1 Minor,s2 Major,s3 |
||||||
|
Purchase insurance,d1 Do not purchase insurance,d2 |
10,000 0 |
10,000 100,000 |
10,000 200,000 |
|||
|
Probabilities 0 |
96 0 |
03 0 |
01 |
|||
a. Using the expected value approach, what decision do you recommend?
b. Assume that you found the following indifference probabilities for the decision alternatives. What decision would you recommend?
|
Cost Indifference Probability |
|
10,000 p=0.99 100,000 p=0.60 |
c. Do you favor using expected value or expected utility approach for this decision problem? Why?
Question 2. (4 + 4 + 2 = 10 marks)
The showtime.xlsx dataset shows the gross revenue, expenditure in television and newspaper advertisements (in $ thousands) for some movie theaters in USA for a given year. Using the data:
a. Check whether the true mean of spending in television advertisements is greater than newspaper advertisements. In doing so, indicate the appropriate null and alternative hypothesis.
b. Develop a regression model which helps demonstrate how the advertisement expenditure explains the gross revenue. After estimating this linear regression model, report the estimated coefficients of the key variables of interest and interpret their relationship with the outcome variable.
c. What are the managerial implications of these results?
Question 3. (4 + 3 + 4 + 6 = 17 marks)
The yelp.csv file contains reviews for an Italian restaurant. Utilizing this document:
a. Normalize/preprocess the terms by removing punctuations, numbers, extra spaces and stop words, converting text to lowercase, and stemming. Show the third document of the reviews before and after the normalizing process.
b. Generate a frequency bar chart. Identify the three most common terms.
c. Generate a word cloud to identify 20 widely mentioned terms. What is the most mentioned word?
d. Using Jaccards distance to compute the dissimilarity between observations, apply hier- archical clustering employing the complete linkage method to yield four clusters from the binary document-term matrix using the three most common terms from the review. Present a table showing the number of mentions of the three most common terms for each cluster and the proportion of documents assigned to each cluster.
2023-06-07