ITEC-620 Business Insights through Analytics Assignment

QUESTION 1(2 pts.) Which of the following is an example of prescriptive analytics?

A) Predicting the amount of rainfall in Maryland next year
B) Classifying an investment as high-risk or low-risk
C) Selecting the best supplier based on projected performance
D) Identifying the relationship between labor, capital, and revenue

QUESTION 2(2 pts.) Which of the following is a property of effectively-designed tables:

A) Only one type of variable is included
B) Text in the body of the table is left-aligned
C) Multiple fonts are used
D) The data-ink ratio is low

QUESTION 3(2 pts.) Which of the following methods will provide the probability of an outcome?

A) k-nearest neighbors
B) Exponential smoothing
C) Classification tree
D) Logistic regression

QUESTION 4(2 pts.) Which of the following statements about ensemble methods is true?

A) Bagging involves creating many large trees
B) Random trees are useful when we have very few independent variables
C) One large regression tree gives better predictions than aggregating many small regression trees
D) Bagging and random trees involve creating many weaker trees

QUESTION 5(2 pts.) What does the following line of R code do:training <- sample(1:1000, 600)

A) Stores the integers from 1 to 600 in order in an object called training
B) Stores 600 different randomly chosen rows from a data set with 1000 rows in an object called training
C) Stores 1000 randomly chosen integers from 1 to 600 in an object called training
D) Stores 600 different randomly chosen integers from 1 to 1000 in an object called training

QUESTION 6(2 pts.) What does the following line of R code do:ggplot(patients,aes(y=pulse,x=age)) + geom_point()

A) Creates a line chart of a patient’s pulse over time
B) Creates a scatterplot of patients’ age and pulse
C) Creates a bar chart of patients’ average pulse, grouped by age
D) Creates a histogram of patients’ pulses, color-coded by age

QUESTION 7(2 pts.) Which of the following is a common reason for a chart to have a high lie factor:

A) Not enough data points are used
B) The y-axis begins at a high number
C) Too many variables are included
D) Data labels are duplicated

QUESTION 8(2 pts.) Expected value is NOT an appropriate method to use for decision making if:

A) One of the possible outcomes would be catastrophic
B) None of the outcomes are very unlikely
C) The decision being analyzed is made frequently
D) All of the impacts of the decision are minor

The following decision tree reflects an energy company’s decision about whether or not to frack a site that they believe contains a substantial quantity of natural gas. If the quantity of natural gas is high enough, they will build new infrastructure in the area to support their operations, which increases the likelihood of negative publicity after they begin.

The only information missing from the decision tree is the probability of the quantity of natural gas being high, for which they have developed a logistic regression model.

a) (2 pts.) Using the logistic regression model, the company’s analytics team has estimated a 0.4 probability that the site will have a high quantity of natural gas. Solve the decision tree, showing the expected value associated with each node of the tree. Based on this decision tree, should they choose to frack the site? 

b) (2 pts.) The energy company would like you to do sensitivity analysis on the probability of the quantity of natural gas being high. Briefly (1-2 sentences), what are the main conclusions of that sensitivity analysis?

The following logistic regression output is from a model used to predict whether or not a country’s average life expectancy will exceed 75 years, based on other demographic data.

Coefficients:

                Estimate Std. Error z value Pr(>|z|) 

(Intercept)    1.026e+01  1.404e+01   0.731   0.4647 

Population     3.652e-04  7.588e-04   0.272   0.7823 

Per Capita GDP 2.756e-03  1.239e-03   2.971   0.0030 ** 

Crime Index   -1.878e+00  0.459e+00  -2.119   0.0342 *

---

Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

a) (2 pts.) Which of the independent variables are likely to be useful for estimating the probability that average life expectancy will exceed 75 years?

b) (2 pts.) According to this model, does an increase in Per Capita GDP lead to a higher or lower probability of life expectancy exceeding 75 years?

The classification tree below is used to predict whether or not a charity’s request for donations by mail will be successful (indicated by a 1). The following independent variables are used:

previous_donor: an indicator variable equal to 1 if the person has given to this charity before, and 0 if not
months_since_last_donation: for previous donors, the number of months since their last donation
income: the average household income of the person’s neighborhood

a) (2 pts.) When making binary classifications, does the tree classify the following person as a 0 or a 1?

previous_donor = 1
months_since_last_donation = 6
income: = $127,500

b) (2 pts.) Briefly (1-2 sentences) explain the logic that this tree is using to make classifications.

c) (2 pts.) The following classification confusion matrix shows the results of this tree’s classifications on the test set:

What is the error rate when the tree predicts TRUE?