Publicação
Forecasting demand in the clothing industry
| Resumo: | For many clothing companies the range of goods sold is renewed twice a year. Each new collection includes a large number of new items which have a short and well defined selling period corresponding to one selling season (20-30 weeks). The absence of past sales data, changes in fashion and product design causes difficulties in forecasting demand accurately. Thus, the predominant factors in this environment are the difficulty in obtaining accurate demand forecasts and the short selling season for goods. An inventory management system designed to operate in this context is therefore constrained by the fact that demand for many goods will not generally continue into the future. Using data for one particular company, the accuracy of demand forecasts obtained by traditional profile methods is analysed and a new approach to demand forecasting using artificial neural networks is presented. Some of the main questions concerning the implementation of neural network models are discussed. |
|---|---|
| Autores principais: | Rodrigues, Eduardo J. |
| Outros Autores: | Figueiredo, Manuel |
| Assunto: | Forecasting demand Clothing industry Artificial neural networks |
| Ano: | 2013 |
| País: | Portugal |
| Tipo de documento: | comunicação em conferência |
| Tipo de acesso: | acesso aberto |
| Instituição associada: | Universidade do Minho |
| Idioma: | inglês |
| Origem: | RepositóriUM - Universidade do Minho |
| Resumo: | For many clothing companies the range of goods sold is renewed twice a year. Each new collection includes a large number of new items which have a short and well defined selling period corresponding to one selling season (20-30 weeks). The absence of past sales data, changes in fashion and product design causes difficulties in forecasting demand accurately. Thus, the predominant factors in this environment are the difficulty in obtaining accurate demand forecasts and the short selling season for goods. An inventory management system designed to operate in this context is therefore constrained by the fact that demand for many goods will not generally continue into the future. Using data for one particular company, the accuracy of demand forecasts obtained by traditional profile methods is analysed and a new approach to demand forecasting using artificial neural networks is presented. Some of the main questions concerning the implementation of neural network models are discussed. |
|---|