Question 2. (10 pts.) You have a training set, a test set and a learning algorithm (for instance a logistic regressor). Choose either True or False and briefly explain each why. (a) - T/F A zero training set error indicates good generalization performance. (b) - T/F A system that has higher test set error compared to its training set error has overfit to the training set. (c) T/F As the number of features increases, the risk of overfitting generally increases. (d) T/F As the number of training samples increases, the risk of overfitting generally decreases. (e) - T/F More complex models with larger number of parameters may fit the training data well, but they are more likely to overfit compared to smaller models