In the k-Nearest Neighbors method for classification, you set k = 1 to 20 and XLMiner reported the best k = 10. The table below gives the 10 nearest neighbors in the training set for a new observation and the class membership for each neighbor. 1 2 Neighbor 3 4 5 7 8 a 10 Class 0 0 1 0 1 0 1 0 0 Answer the following questions based on the above information. Question 53 (4 points) If the cutoff value is 0.50, the new observation should be classified as Cannot be determined; more information is needed. Class 1 Class o Question 54 (4 points) For the new observation, the probability of being in Class 1 = A Use 2-decimal accuracy for the final answer, e.g., 0.12, when necessary. Question 55 (4 points) In the data partitioning procedure, if a rare event is involved in classifying a categorical outcome, then should be used for the training set. overfitting oversampling standard partitioning standard sampling Question 56 (4 points) The best k achieves the smallest on the validation set. class 0 error rate class 1 error rate overall error rate root mean squared error Previous Page Next Page Page 39