Choose the most accurate and complete answer below: a. The objective of evaluating a model's performance on hold-out (test) set is to approximate the model's training performance. b. The objective of evaluating a model's performance on hold-out (test) set is to approximate the model's performance once it is deployed in practice to predict the outcomes of new cases. c. No answer is correct. 2. Select the most complete and correct answer. a. A given model's training accuracy is typically higher than the model's test accuracy. b. A given model's training accuracy is typically lower than the model's test accuracy. c. A model's training accuracy is always the same as the model's test accuracy d. When comparing between the performances of two alternative classification models, the model which has higher training accuracy is likely to be a better predictive model since it fits the training data better. e. No answers is correct