Forecasting methods using fuzzy concepts

In this paper, fuzzy concepts are applied in forecasting product price and sales in the semiconductor industry which is often conceived as a highly dynamic environment. First, two fuzzy forecasting methods including fuzzy interpolation (FI) and fuzzy linear regression (FLR) are developed and discussed. Forecasts generated by these methods are fuzzy-valued. Next, the subjective beliefs about whether the industry is booming or slumping, and the speed at which this change in prosperity takes place during a given period are also considered. Two subjective functions are defined and used to adjust fuzzy forecasts. Practically, fuzzy forecasts are incorporated with fuzzy programming like fuzzy linear programming (FLP) or fuzzy nonlinear programming (FNP) for mid-term or long-term planning. Advantages over traditional methods are shown in our examples. (C) 1999 Elsevier Science B.V. All rights reserved.