A key to success in the worldwide competition, for textile-ganmnt companies, is the lean produciion. To produce in time suitable items at sufficient quantity, it is necessary to build a sales forecasting system adapted to an uncertain environment of the textile background. This system can be characterized using noisy data and multiple explanatory variables (controlled, available or not) related to the sales behavior. The proposed model is a hybrid neural model (HNCCX). It uses neural networks ability to model the nonlinear influences of explanatory variables, and considers the seasonality factors. The architecture and the learning procedures of this model are described in the paper. This model is essentially adapted to a mean-term forecasting horizon, and to short and discontinuous time series, i.e. when the product sales occur in some periods of a year only (major cases of textile items). To evaluate perfomances, the proposed prediction model is compared with some existing ones and applied to real data of textile items from an important French ready-to-wear distributor. Further works are also introduced in order to improve the quality of the proposed model.
Textile-apparel industry, Sales forecasting, Artificial neural network, Production and distribution management.
All articles published by Studies in Informatics and Control (SIC) are published in accordance with the terms and conditions of the Creative Commons Attribution License (CC BY-NC-SA 4.0).