Sales forcasting with atrificial neural network at an apparel retailing company

In this century, businesses have to decide the best decision for resisting undercondition of market. The factor of envirement and number of the competitors areincreasing vague for taking decision. If the businesses estimate demand close to real,they can locate themselves a good position and take strategic decisions which providesuccess. At Apparel retailing, qualitative, quantative methods and hibrit of both ofthem could be used for sales forecating. Because of trend and seasonality and shortproduct life cycle for long and medium term sales forcast qualitative methods can beused, for short term sale forecast quantative methods can be used. Using salesforecast for qualitative method with neural network provides more accurate result fornonlinear demands.On this work, NARX (nonlinear autoregressive exogenous) model which is a kind ofneural network is determined the best way for 2011 sales forecasting. At the firststage, mentioned apparel retailing, at the second stage mentioned qualitative,quantative and hibit of these methods for apparel retailing sector. At the third stage, mentioned neural network structure and at the final stage, metioned NARX methodand application of this method on an apparel retailing.Key words: Artificial neural network, NARX, sale forecasting, apparel