Answer with brief justification a. A rule-based classifier is built through indirect method from a decision tree. The decision tree turns out to be overfit on training data. What will be the impact on rule-based classifier? (3 marks) b. The following training dataset provides income (can be high-H, medium-M, or low-L) and education level (can be high-H or low-L) of customers who bought a smart watch. We intend to construct decision tree to predict who buys smart watch. Identify the attribute for root node using the Gain Ratio (5 marks) Income H H L M M L H H M H L L M Education H H H H L L L L H IL H Bought N Y Y Y Y N N Y Y Y N Smart watch H L L Y Y Y