Using neural networks for sales forecast in retail industry

Abstract

This paper explores the use of artificial neural networks in sales forecasts in retailing industry. Historical data is used to make forecasts of product sales in the retailing industry. Weekly sales data from individual products were used to generate models. Results sugest that models based upon neural networks can be adequate to individual product sales forecast. Models obtained with the proposed methodology are able to predict product sales in the short term more accurately then non adjusted naïve techniques and linear regression, more frequently used.