Artificial neural networks for demand forecasting: Application using Moroccan supermarket data

Abstract

The accuracy of sales forecasts in a supply chain is certainly an important key to competitiveness. Because, for any member of the supply chain system, having a clear vision of the future demand affects his planning, his performance so his profit. In the first study of this work, various Artificial Neural Network models were presented and utilized to predict demand of a costumer's product. The training and validating data are provided from a known supermarket in Morocco. In a previous study, the results indicated that the best neural network structure for demand forecasting is the Multi Layer Perceptron, which is by the way, the most commonly used model in the literature. This work focuses on finding the optimal Multi Layer Perceptron structure for demand forecasting. We also present a review of selected works done in the application of game theory and neural networks in the context of management science. The main contribution of our work is the use of neural networks in order to predict the consumer's demand and implement this demand forecasting in a two-echelon supply chain with a game theoretic approach.