Solving the one-shot decision theory based newsvendor models

摘要

The one-shot decision theory (OSDT)has been utilized to solve newsvendor models for innovative products. In fact, OSDT based newsvendor models are a special kind of bilevel programming problems. The existing approaches may not be applicable for solving these models because they include nonconvex nonsmooth lower level optimization problems. In this paper, we translate these models into conventional optimization models with considering the characteristics of the optimal solutions, which effectively handle the difficulties of the original newsvendor models. The theoretical results are given and a numerical example is used to show the effectiveness of the proposed methods.