This paper is concerned with multi-objective linear programming problems in which the objective functions can be partially ranked. We represent the set of admissible weight vectors by a system of linear constraints and solve for the policy most likely to be optimum. If each admissible weight vector has the same probability of being correct, the optimum policy maximizes the hypervolume of the polytope of weight vectors having this policy as a solution. The proposed algorithm requires the enumeration of the subset of admissible efficient solutions of a multi-objective linear program. For each admissible solution, we estimate the ratio of the volumes of the corresponding polytope of weight vectors and the polytope of all admissible weight vectors. An algorithm is outlined for numerical integration using the Monte Carlo method. The model is extended to the case where several objectives are expressed as linear constraints with multiple parameter vectors and there is uncertainty about the weighting of these parameters. A numerical example is provided.