When testing, instance based learning requires a distance comparison to: each example in the training set each example in the test set a decision tree Question 4 Clustering results in : classifier sets of unlabeled examples called clusters association rules labels for all test data