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 prior to implementation, vary from period to period. These vectors,...