Approximate Solutions of a Dynamic Forecast-Inventory Model

In this paper we consider a dynamic forecast-inventory model with forecast updates, based on the martingale model of forecast evolution. Two types of updates are considered, additive and multiplicative. The formulation of the model results in a dynamic program with multidimensional state space. We derive some characteristics of optimal policies and also develop a computational approach to obtain approximate solutions. The approach is based on simulation and function approximation.