Suppose a bank makes loan decisions using two decision trees, one that uses attributes related to credit history and one that uses other demographic attributes. Each decision tree separately classifies a loan applicant as High Risk or Low Risk. The bank only offers a loan when both decision trees predict Low Risk.

(a) Describe an algorithm for converting this pair of decision trees into a single decision tree that makes the same predictions (that is, it predicts non-risky only when both of the original decision trees would have predicted non-risky).

(b) Let n1 and n2 be the number of leaves in the first and second decision trees, respectively. Provide an upper bound on n, the number of leaves in the single equivalent decision tree, expressed as a function of n1 and n2.