Forecasting the output of integrated circuit industry using a grey model improved by the Bayesian analysis

The production values of the integrated circuit industry has the following attributes, short product life cycle, numerous influencing factors on the market, and rapid changing of technology. These features obstruct the precision of forecasting the outputs of integrated circuit industry using the traditional statistical methods. The grey forecast model can obviously conquer these difficulties with a small sample set and ambiguity of available information. This study evaluates original and Bayesian grey forecast models for the integrated circuit industry. Bayesian method uses the technique of Markov Chain Monte Carlo to estimate the parameters for grey differential function. The predictive value of integrated circuit in Taiwan was evaluated along with mean absolute percentage error. Various parameters and efficiency of three forecast models were compared and summary outcomes were reported. Meanwhile, the Bayesian grey model was the most accurate one among these models.