Evaluating Mixed-Integer Programming Models over Multiple Right-hand Sides

A critical measure of model quality for a mixed-integer program (MIP) is the difference, or gap, between its optimal objective value and that of its linear programming relaxation. Although in many contexts, only an approximation of the right-hand side(s) is available, there is no consensus on appropriate measures for MIP model quality over multiple right-hand sides. In this paper, we provide model formulations for the expectation and extrema of absolute and relative MIP gap functions over finite discrete sets.

Article

Download

View Evaluating Mixed-Integer Programming Models over Multiple Right-hand Sides