Application of gray ANN model in products demand forecasting for the Fabless

Fabless market demand is a multi-factors and complex time-varying nonlinear link with a high degree of uncertainty, this uncertainty makes it difficult to predict. In this paper, a grey artificial neural network model is put forward, and the model is combined with grey system theory and artificial neural network theory. Using annual sales of the certain type chips between 2002 and 2009 of a Fabless company in shanghai, to simulate and predict the last five years´ wafer annual sales of the certain types of chips the company, the results show that the maximum relative error between predictions and actual value is 0.28%, indicates that the model predictions approach to the actual values considerably. When the model is used to forecast the IC product and describe the development and changes of the product accurately, it will be helpful for the enterprise to make decisions.