CHAPTER 9: PREDICTIVE DATA MINING PART A: k-NN for Classification You used the k-Nearest Neighbors method for classification (e.g., whether a customer should be classified as Default or No Default on a loan). You set = 1 to 20 and the data mining software reported the best_k= 9. The table below gives the 9 nearest neighbors in the training set for a new observation and the class membership for each neighbor. N w - wn Neighbor 1 ~ \\o Class 1 0 1 1 0 1 Answer the following questions based on the above information. Question 14: The best / achieves the smallest on the validation set. a. class | error rate b. class 0 error rate c. overall error rate d. root mean squared error Question 15: For the new observation, the probability of being in Class 1 = a. 0.40 b. 0.50 c. 0.44 d. 0.56 Question 16: If the cutoff value is 0.50, the new observation should be classified as a. Class 0 b. Class 1 c. cannot tell; more information is needed. Question 17: If the cutoff value is 0.60, the new observation should be classified as a. Class | b. Class 0 c. cannot tell, more information is needed