We study dynamic decision making under uncertainty when, at each period, a decision maker implements a solution to a combinatorial optimization problem. The objective coefficient vectors of said problem, which are unobserved ...