Inverse Optimization: A Case Study in Supply Chain Cost Estimation

Abstract

Determining unknown parameters of an optimization model such that a given feasible solution becomes optimal is the focus of the quickly emerging discipline of inverse optimization in operations research. Inverse optimization infers model coefficients from observed decisions, in contrast to traditional optimization, which seeks the optimal solution for the given know parameters. This Paper demonstrate the implementation of inverse optimization through a case study in supply chain cost estimation. The inverse problem is developed to extract transportation cost coefficients from observable shipping patterns using a linear programming transportation model. The formulation in mathematics, the method of solving it, and the computing example are presented. The paper emphasises how inverse optimization is useful in decision sciences, logistics, energy systems, and economics.