D Question 3 3 pts Suppose we fit a decision tree classifier to a dataset and obtain a training accuracy of 0.95 and a cross-validation accuracy of 0.75. If we apply pruning on this decision tree, which of the following would be the ideal scenario? O Training accuracy and cross-validation accuracy remains unchanged Training accuracy increases (say 0.98) but cross-validation accuracy remains unchanged (0.75) Training accuracy reduces marginally (say 0.9) and cross-validation accuracy increases (say 0.85) O Training accuracy decreases marginally (say 0.9) and cross-validation accuracy decreases marginally (say 0.7) 3 pts