,ISYE 6501 Introduction to Analytics Modeling
Comprehensive Practice Exam: Weeks 1–11
Week 1: Introduction and Classification
1. Which type of analytics question asks "What happened?"
A. Predictive
B. Prescriptive
C. Descriptive
D. Diagnostic
Correct Answer: C
Rationale: Descriptive analytics answers questions about what has already
occurred, such as "Which customers are most alike in their buying patterns?" or
"When did equipment reliability drop below acceptable levels?" . Predictive
analytics forecasts future outcomes, while prescriptive analytics recommends
actions.
2. A soft classifier is used when:
A. All data points can be perfectly separated
B. There is no line that perfectly separates all labeled examples
C. The dataset contains only two classes
D. The classifier uses a nonlinear kernel
Correct Answer: B
Rationale: Soft classifiers minimize the number of mistakes when perfect
separation is impossible. They trade off reducing classification errors against
enlarging the margin between classes .
, 3. In a classification problem, what does a decision boundary that is nearly
parallel to the vertical axis indicate?
A. The horizontal attribute is most important
B. The vertical attribute is most important
C. Both attributes are equally important
D. Neither attribute is useful
Correct Answer: A
Rationale: When the classifier boundary is almost parallel to the vertical axis,
classification decisions depend primarily on the horizontal attribute (x-axis). The
vertical attribute has little influence on the classification .
4. Which of the following is an example of categorical data?
A. Temperature in degrees Fahrenheit
B. Annual income in dollars
C. ZIP code
D. Age in years
Correct Answer: C
Rationale: Categorical data consists of numbers without inherent meaning (like
ZIP codes) or non-numeric categories (like hair color). Quantitative data has
arithmetic meaning where higher means more and lower means less .
5. What is the defining characteristic of supervised learning?
A. The algorithm discovers hidden patterns without labels
B. Both features and response variables are known
C. Only the response variable is used
D. The data is always time-series
Correct Answer: B
Comprehensive Practice Exam: Weeks 1–11
Week 1: Introduction and Classification
1. Which type of analytics question asks "What happened?"
A. Predictive
B. Prescriptive
C. Descriptive
D. Diagnostic
Correct Answer: C
Rationale: Descriptive analytics answers questions about what has already
occurred, such as "Which customers are most alike in their buying patterns?" or
"When did equipment reliability drop below acceptable levels?" . Predictive
analytics forecasts future outcomes, while prescriptive analytics recommends
actions.
2. A soft classifier is used when:
A. All data points can be perfectly separated
B. There is no line that perfectly separates all labeled examples
C. The dataset contains only two classes
D. The classifier uses a nonlinear kernel
Correct Answer: B
Rationale: Soft classifiers minimize the number of mistakes when perfect
separation is impossible. They trade off reducing classification errors against
enlarging the margin between classes .
, 3. In a classification problem, what does a decision boundary that is nearly
parallel to the vertical axis indicate?
A. The horizontal attribute is most important
B. The vertical attribute is most important
C. Both attributes are equally important
D. Neither attribute is useful
Correct Answer: A
Rationale: When the classifier boundary is almost parallel to the vertical axis,
classification decisions depend primarily on the horizontal attribute (x-axis). The
vertical attribute has little influence on the classification .
4. Which of the following is an example of categorical data?
A. Temperature in degrees Fahrenheit
B. Annual income in dollars
C. ZIP code
D. Age in years
Correct Answer: C
Rationale: Categorical data consists of numbers without inherent meaning (like
ZIP codes) or non-numeric categories (like hair color). Quantitative data has
arithmetic meaning where higher means more and lower means less .
5. What is the defining characteristic of supervised learning?
A. The algorithm discovers hidden patterns without labels
B. Both features and response variables are known
C. Only the response variable is used
D. The data is always time-series
Correct Answer: B
