Download ISYE 6501 Midterm 1 _ Intro Analytics Modeling - ISYE-6501_O01-OAN_O01_MSA and more Exams Ecosystem Modelling in PDF only on Docsity!
ISYE 6501 Midterm 1
- Due Mar 15 at 2am
- Points 100
- Questions 45
- Available until Mar 15 at 2am
- Time Limit 95 Minutes
Instructions
This quiz was locked Mar 15 at 2am. This element is a more accessible alternative to drag & drop reordering. Press Enter or Space to move this question. Question 1 9 / 13 pts
- For each of the 13 models/methods, select the choice that includes the category of question it is commonly used for. For models/methods that have more than one correct category, the one it is most commonly used for; for models/methods that have no correct category listed, select "None". i. ARIMA Response prediction ii. CART Classification and Response prediction iii. Cross validation Validation iv. CUSUM None of the other choices v. Exponential smoothing Response prediction vi. GARCH Variance estimation vii. kmeans Classification viii. k-nearest-neighbor Clustering ix. Linear regression Validation x. Logistic regression Classification and Response prediction xi. Principal component analysis Validation xii. Random forest Classification and Response prediction xiii. Support vector machine Classification
Answer 1: Classification Clustering Correct! Response prediction Validation Variance estimation None of the other choices Answer 2: Correct! Classification and Response prediction Clustering Validation Variance estimation None of the other choices Answer 3: Classification and Response prediction
Variance estimation None of the other choices Answer 6: Classification Clustering Respnse prediction Validation Correct! Variance estimation None of the other choices Answer 7: You Answered Classification Correct Answer Clustering Response prediction Validation Variance estimation
None of the other choices Answer 8: Correct Answer Classification and Response prediction You Answered Clustering Validation Variance estimation None of the other choices Answer 9: Classification Clustering Correct! Response prediction Validation Variance estimation None of the other choices Answer 10: Correct! Classification and Response prediction
None of the other choices Answer 13: Correct! Classification Clustering Response prediction Validation Variance estimation None of the other choices his element is a more accessible alternative to drag & drop reordering. Press Enter or Space to move this question. Question 2 3 / 3 pts
- For each of the following models, specify whether it is designed for use with attribute/feature data or time-series data: a. Exponential smoothing [ Select ] ["Attribute/feature data", "Time series data"] b. ARIMA [ Select ] ["Attribute/feature data", "Time series data"] c. k-means [ Select ] ["Attribute/feature data", "Time series data"] d. Principal component analysis [ Select ] ["Attribute/feature data", "Time series data"] e. Linear regression [ Select ] ["Attribute/feature data", "Time series data"] f. k-nearest-neighbor Attribute/feature data g. Random forest [ Select ] ["Attribute/feature data", "Time series data"] h. CUSUM [ Select ] ["Attribute/feature data", "Time series data"] i. Logistic regression [ Select ] ["Attribute/feature data", "Time series data"] j. Support vector machine [ Select ] ["Attribute/feature data", "Time series data"] k. GARCH [ Select ] ["Attribute/feature data", "Time series data"] Answer 1:
Attribute/feature data Correct! Time series data Answer 2: Attribute/feature data Correct! Time series data Answer 3: Correct! Attribute/feature data Time series data Answer 4: Correct! Attribute/feature data Time series data Answer 5: Correct! Attribute/feature data Time series data Answer 6: Correct! Attribute/feature data Time series data
Figures A and B show the training data for a soft classification problem, using two predictors (x 1 and x 2 ) to separate between black and white points. The dashed lines are the classifiers found using SVM. Figure A uses a linear kernel, and Figure B uses a nonlinear kernel that required fitting 16 parameter values. Figure A
Figure B
Move To... This element is a more accessible alternative to drag & drop reordering. Press Enter or Space
to move this question. INSTRUCTIONS FOR QUESTIONS 3- 11 For each statement in Questions 3-11, select the choice that makes the statement true.
Move To... This element is a more accessible alternative to drag & drop reordering. Press Enter or Space
to move this question. Question 3 0.6 / 0.6 pts Figure A's classifier IS NOT based on the values of both x 1 and x 2. Answer 1: IS Correct! IS NOT
Move To... This element is a more accessible alternative to drag & drop reordering. Press Enter or Space
to move this question. Question 7 0.6 / 0.6 pts Figure A DOES NOT SHOW that the black point (7.2,1.4) is an outlier. Answer 1: Correct! DOES NOT SHOW SHOWS
Move To... This element is a more accessible alternative to drag & drop reordering. Press Enter or Space
to move this question. Question 8 0.75 / 0.75 pts Figure B's classifier has a NARROWER margin in the training data than Figure A's classifier. Answer 1: Correct! NARROWER WIDER
Move To... This element is a more accessible alternative to drag & drop reordering. Press Enter or Space
to move this question. Question 9 0 / 0.75 pts Figure B's classifier would probably perform BETTER on test data than on training data. Answer 1: BETTER THE SAME Correct! WORSE
Move To... This element is a more accessible alternative to drag & drop reordering. Press Enter or Space
to move this question. Question 10 0.75 / 0.75 pts Figure B's classifier incorrectly classifies EXACTLY 5 black points in the training data. Answer 1: Correct! EXACTLY 5 MORE OR FEWER THAN 5
Move To... This element is a more accessible alternative to drag & drop reordering. Press Enter or Space
to move this question. Question 11 0.75 / 0.75 pts Figure B DOES NOT SHOW that the black point (7.2,1.4) is an outlier. Answer 1: Correct! DOES NOT SHOW SHOWS
Move To... This element is a more accessible alternative to drag & drop reordering. Press Enter or Space
to move this question. Question 12 2.25 / 3 pts For each of the following statements, select the correct choice to make the statement true. i. A new point at (4.9, 4) would be classified as WHITE by Figure A's classifier. ii. A new point at (4.9, 4) would be classified as BLACK by Figure B's classifier. iii. A new pint at (4.9, 4) would be classified as WHITE by a k-nearest neighbor algorithm where k=5. iv. In Figure A, if the training data had 1000 more white points to the right of the classifier, a 1000-nearest-neighbor algorithm would classify a new point at (4.9, ,4) as WHITE. Answer 1:
Question 13 2 / 3 pts For each statement, select the choice that makes the statement correct. i. Decreasing the value of C could INCREASE the margin. ii. Requiring a larger margin could INCREASE the number of classification errors in the training set. iii. Decreasing the value of C could REDUCE the number of classification errors in the training set. Answer 1: You Answered INCREASE Correct Answer REDUCE Answer 2: Correct! INCREASE REDUCE Answer 3: INCREASE Correct! REDUCE
Move To... This element is a more accessible alternative to drag & drop reordering. Press Enter or Space
to move this question. INFORMATION FOR QUESTIONS 14- 15 For each of the following statements about the hard classification SVM model, select the choice that makes the statement true.
Move To... This element is a more accessible alternative to drag & drop reordering. Press Enter or Space
to move this question. Question 14 2 / 2 pts Moving the classifier from the location that has equal margin on both sides is more likely to result in i. MORE classification errors in the validation data. ii. MORE classification errors in the test data. Answer 1: FEWER Correct! MORE Answer 2: FEWER Correct! MORE
Move To... This element is a more accessible alternative to drag & drop reordering. Press Enter or Space
to move this question. Question 15 1 / 1 pts It might be desirable to not put the classifier in a location that has equal margin on both sides when the costs of misclassifying the two types of points are SIGNIFICANTLY DIFFERENT. Answer 1: Correct! SIGNIFICANTLY DIFFERENT VERY SIMILAR
Move To... This element is a more accessible alternative to drag & drop reordering. Press Enter or Space
to move this question. INSTRUCTIONS FOR QUESTIONS 16- 17
For each of the statements in Questions 16-17, select the choice that makes the statement true.
Move To... This element is a more accessible alternative to drag & drop reordering. Press Enter or Space
to move this question. Question 16 0 / 3 pts Which of these models would you expect to perform worst on a test data set? Correct Answer Model 1, because it has much lower adjusted R-squared. Model 2, because it's the simplest of those with a high adjusted R-squared. Model 4, because its adjusted R-squared is only slightly lower than Model 5 and uses one fewer predictor. Model 5, because it has the highest adjusted R-squared. You Answered Model 7, because it uses the most predictors. One of Models 2,3,4,5,6,7, but it's hard to be sure which because their adjusted R-squared values are so close to each other.
Move To... This element is a more accessible alternative to drag & drop reordering. Press Enter or Space
to move this question. Question 17 3 / 3 pts
