A sales forecasting model for manufacturer in hybrid supply chain

In view of the characteristics of sales forecasting of manufacturer in hybrid supply chain (HSC), an extended recursive forgetting factor forecasting (ERFFF) model was given in this paper. The coefficient vector in this model is altered dynamically with the actual data, and the optimum structural parameters are adjusted according to the Theil U statistic, thus the best forecasting value is acquired under the optimal state. Through a practical case, it was proven the greater accuracy and applied value of this model.