According to the above analysis, neither the time series nor the exponential smoothing
methods are suitable for the IC industry. Consequently, the Grey model (GM) is applied
herein to predict future trends in the global IC industry. The GM has the following
advantages: (a) It can be used in situations with relatively with limited data down to as little
as four observations. (b) Just a few discrete data are sufficient to characterize an unknown
system. (c) It is suitable for forecasting in competitive environments where decision-makers
only have access to limited historical data. Moreover, three residual modification models
were applied to enhance the GM models.