Academic Integrity: tutoring, explanations, and feedback — we don’t complete graded work or submit on a student’s behalf.

Problem 5 (10 points + Bonus 3 points) Each of the following statements can be T

ID: 375764 • Letter: P

Question

Problem 5 (10 points + Bonus 3 points) Each of the following statements can be TRUE or FALSE. Mark In the . you think it is TRUE, and with an "P" if you think it is FALSE. heed to show prooedures or explanations). t y ealcuiation (1). Bagging trees method is essentially a special case of random forests method (25 (2). ( ) In performing LASSO, a smaller value of the tuning paran eter tends to yield (3). () Logistie regression and LDA give both linear classification boundaries. (4). ( Input variables need to be centered before implementing ridge regression. (2.5 pointa) (5). Bonus question: Cross-validation error is a good estimate for training error. (3 points) smaller number of input variables to be automatically selected. (2.5 points) (2.5 points) points) 10

Explanation / Answer

1. false. Random forest method is an improvement over bagging method.

2. True. controls amount of regularization.

3. True. Both Logistic Regression and Linear Discriminant Analysis are considered appropriate for the development of linear classification models.

4. True: it is assumed that input variables are centred and do not include any intercept while implementing ridge regression.

5. False:

5.