ISYE 6501 INTRODUCTION TO
ANALYTICS MODELING FINAL EXAM
QUESTIONS AND ANSWERS
1. In Support Vector Machines (SVM), what is the primary purpose of the ‘C’ parameter?
A. To control the trade-off between maximizing the margin and minimizing classification
errors.
B. To determine the number of support vectors used in the final model.
C. To define the dimensionality of the feature space in a radial kernel.
D. To calculate the distance between the hyperplane and the nearest data point.
Answer: A
Conceptual Explanation: The C parameter is a regularization parameter that balances the
margin size against the training error. A large C penalizes errors heavily, leading to a
smaller margin.
2. When using K-Nearest Neighbors (KNN), what effect does increasing ‘k’ typically have on
the model?
A. It increases the variance and makes the model more sensitive to outliers.
B. It decreases the bias but significantly increases the computational complexity.
,C. It ensures that the model will always converge to a global minimum.
D. It smooths the decision boundary, reducing variance but potentially increasing bias.
Answer: D
Conceptual Explanation: A larger k considers more neighbors, which leads to a smoother
decision boundary (lower variance) but may ignore local patterns (higher bias).
3. Which of the following techniques is most appropriate for identifying seasonal patterns
and trends simultaneously in a time series?
A. Simple Moving Average
B. Principal Component Analysis
C. K-Means Clustering
D. Holt-Winters Exponential Smoothing
Answer: D
Conceptual Explanation: Holt-Winters is designed specifically to handle time series data
with both trend (Holt) and seasonality (Winters).
4. In the context of K-means clustering, how is the ‘Elbow Method’ used?
A. To determine the optimal distance metric between data points.
B. To visualize the separation between different hierarchical clusters.
C. To calculate the probability that a point belongs to a specific cluster.
, D. To identify the point where adding more clusters no longer significantly reduces Within-
Cluster Sum of Squares (WCSS).
Answer: D
Conceptual Explanation: The Elbow Method plots WCSS against the number of clusters;
the ‘elbow’ point suggests the optimal k where gains in variance reduction diminish.
5. What is the key difference between Lasso (L1) and Ridge (L2) regression?
A. Ridge can shrink coefficients exactly to zero, while Lasso cannot.
B. Lasso is used for classification, while Ridge is used for regression.
C. Lasso can perform feature selection by shrinking coefficients to zero, while Ridge only
reduces them.
D. Ridge regression requires a higher number of observations than Lasso.
Answer: C
Conceptual Explanation: Lasso uses an L1 penalty which creates sparsity, effectively
performing feature selection. Ridge uses an L2 penalty which shrinks coefficients but rarely
to zero.
6. If you are monitoring a process to detect a sudden change in mean, which model is most
suitable?
A. Logistic Regression
B. Factor Analysis
ANALYTICS MODELING FINAL EXAM
QUESTIONS AND ANSWERS
1. In Support Vector Machines (SVM), what is the primary purpose of the ‘C’ parameter?
A. To control the trade-off between maximizing the margin and minimizing classification
errors.
B. To determine the number of support vectors used in the final model.
C. To define the dimensionality of the feature space in a radial kernel.
D. To calculate the distance between the hyperplane and the nearest data point.
Answer: A
Conceptual Explanation: The C parameter is a regularization parameter that balances the
margin size against the training error. A large C penalizes errors heavily, leading to a
smaller margin.
2. When using K-Nearest Neighbors (KNN), what effect does increasing ‘k’ typically have on
the model?
A. It increases the variance and makes the model more sensitive to outliers.
B. It decreases the bias but significantly increases the computational complexity.
,C. It ensures that the model will always converge to a global minimum.
D. It smooths the decision boundary, reducing variance but potentially increasing bias.
Answer: D
Conceptual Explanation: A larger k considers more neighbors, which leads to a smoother
decision boundary (lower variance) but may ignore local patterns (higher bias).
3. Which of the following techniques is most appropriate for identifying seasonal patterns
and trends simultaneously in a time series?
A. Simple Moving Average
B. Principal Component Analysis
C. K-Means Clustering
D. Holt-Winters Exponential Smoothing
Answer: D
Conceptual Explanation: Holt-Winters is designed specifically to handle time series data
with both trend (Holt) and seasonality (Winters).
4. In the context of K-means clustering, how is the ‘Elbow Method’ used?
A. To determine the optimal distance metric between data points.
B. To visualize the separation between different hierarchical clusters.
C. To calculate the probability that a point belongs to a specific cluster.
, D. To identify the point where adding more clusters no longer significantly reduces Within-
Cluster Sum of Squares (WCSS).
Answer: D
Conceptual Explanation: The Elbow Method plots WCSS against the number of clusters;
the ‘elbow’ point suggests the optimal k where gains in variance reduction diminish.
5. What is the key difference between Lasso (L1) and Ridge (L2) regression?
A. Ridge can shrink coefficients exactly to zero, while Lasso cannot.
B. Lasso is used for classification, while Ridge is used for regression.
C. Lasso can perform feature selection by shrinking coefficients to zero, while Ridge only
reduces them.
D. Ridge regression requires a higher number of observations than Lasso.
Answer: C
Conceptual Explanation: Lasso uses an L1 penalty which creates sparsity, effectively
performing feature selection. Ridge uses an L2 penalty which shrinks coefficients but rarely
to zero.
6. If you are monitoring a process to detect a sudden change in mean, which model is most
suitable?
A. Logistic Regression
B. Factor Analysis
