1
ISYE6501 / ISYE 6501 Midterm Exam 1 (Latest
Update ) Intro to Analytics
Modeling | Questions & Answers | Grade A |
100% Correct. Georgia Tech
1. Hard classifier: (SVM) Observations can be perfectly separated groups by
classifier line
2. Soft classifier: (SVM) Observations separated as well as possible by the
classifier line
3. When the classifier line is vertical...: (SVM) only x-axis matters in
determining class
4. When the classifier line is horizontal...: (SVM) only y-axis matters in
determining class
5. What is the goal of maximizing the space between support vectors?:
minimize classification error (more generalized model)
6. In terms of the equation, how do we maximize space between support
vectors?: minimize £(aj)^2
, 2
7. In the. equation for a classifier line, if aj is close to 0...: the attribute is
not relevant
8. As »increases...: (SVM) more emphasis placed on maximizing the margin
9. As »decreases...: (SVM) more emphasis placed on minimizing
classification error 10. How do the relationships of »differ when using C
in computation?: (SVM) C has inverse relationships to ». As C increases,
minimization of error is prioritized.
As C decreases, maximization of margin is prioritized
11. What is a weakness of randomly selecting data for
test/train/validation sets?: possible to have clumps of data
12. What is a weakness of the rotation method of selecting data for
test/train/validation sets?: can introduce bias if there are repeating
patterns in the data (i.e., days of the week)
13. If a model requires a bounded range, you need to...: (scaling)
normalize
14. What types of models prefer standardized data?: PCA, clustering
, 3
15. What is a possible effect of using unscaled data in a classifier model?:
introduce bias into the distance metric. Changes in features with larger
ranges will have a disproportionate impact
16. When should you use KNN over SVM?: >2 classes
17. What is the KNN model sensitive to?: noisy predictors and near-0
variance features
18. KNN: A low k-value may _______: overfit
19. KNN: A high k-value may _______: underfit
20. KNN: If there are a large number of (possibly irrelevant) features, a
_______ k-value tends to be better.: larger
21. KNN: what is the benefit of using an odd vs. even number for k?: Odd
number of k avoids ties
22. How is feature importance determined for KNN?: features with
smallest distance measure
23. What values do you need to standardize data?: mean, standard
deviation 24. What do you do when you want a HARD classifier to
ISYE6501 / ISYE 6501 Midterm Exam 1 (Latest
Update ) Intro to Analytics
Modeling | Questions & Answers | Grade A |
100% Correct. Georgia Tech
1. Hard classifier: (SVM) Observations can be perfectly separated groups by
classifier line
2. Soft classifier: (SVM) Observations separated as well as possible by the
classifier line
3. When the classifier line is vertical...: (SVM) only x-axis matters in
determining class
4. When the classifier line is horizontal...: (SVM) only y-axis matters in
determining class
5. What is the goal of maximizing the space between support vectors?:
minimize classification error (more generalized model)
6. In terms of the equation, how do we maximize space between support
vectors?: minimize £(aj)^2
, 2
7. In the. equation for a classifier line, if aj is close to 0...: the attribute is
not relevant
8. As »increases...: (SVM) more emphasis placed on maximizing the margin
9. As »decreases...: (SVM) more emphasis placed on minimizing
classification error 10. How do the relationships of »differ when using C
in computation?: (SVM) C has inverse relationships to ». As C increases,
minimization of error is prioritized.
As C decreases, maximization of margin is prioritized
11. What is a weakness of randomly selecting data for
test/train/validation sets?: possible to have clumps of data
12. What is a weakness of the rotation method of selecting data for
test/train/validation sets?: can introduce bias if there are repeating
patterns in the data (i.e., days of the week)
13. If a model requires a bounded range, you need to...: (scaling)
normalize
14. What types of models prefer standardized data?: PCA, clustering
, 3
15. What is a possible effect of using unscaled data in a classifier model?:
introduce bias into the distance metric. Changes in features with larger
ranges will have a disproportionate impact
16. When should you use KNN over SVM?: >2 classes
17. What is the KNN model sensitive to?: noisy predictors and near-0
variance features
18. KNN: A low k-value may _______: overfit
19. KNN: A high k-value may _______: underfit
20. KNN: If there are a large number of (possibly irrelevant) features, a
_______ k-value tends to be better.: larger
21. KNN: what is the benefit of using an odd vs. even number for k?: Odd
number of k avoids ties
22. How is feature importance determined for KNN?: features with
smallest distance measure
23. What values do you need to standardize data?: mean, standard
deviation 24. What do you do when you want a HARD classifier to
