Deterministic Approximations to Stochastic Production Problems

In this paper we consider deterministic approximations to a nonsequential stochastic production problem with setup costs. We show that, for the most commonly used probability distributions of the demands, the relative error bounds tend to be very small. We also consider a sequential stochastic problem, without setup costs, and discuss its relationship to the nonsequential case. Throughout the paper, we present limited computational results and indicate topics for further research.