| The purpose of this study is to provide a model that can be used to adjust forecasts that are already available. Adjustments in the forecasts of an item may be appropriate when some of the future demands for that item are known in advance. In this dissertation a model is presented which is used to adjust an existing forecast. It is assumed that the already available forecast has been generated without any use of the information known on prior demand.; This research analyzes the components of the advanced demand, namely, the number of orders and their corresponding order size. Particularly, it explores and analyzes the possibility of using the expected number of orders for a future period as the variable to be estimated. The characteristics of the binomial probability distribution, the Poisson probability distribution, and the normal probability distribution along with the basis of Bayes' theorem will be used in the estimation. The expected number of orders is used in determining the adjusted forecast. Several cases are introduced and analyzed. Then, a simulation is used to compare the adjusted forecast with the original one. Finally, a conclusion is drawn based on the analysis of the reduction in the forecasting errors between both models. Results show that the adjusted forecast provides greater accuracy for different probable values of getting an order in advance. |
