What is the training error rate and test error rate of this pruned tree? Would you choose this pruned tree over the original tree for performing classification on unseen instances? Briefly
explain