Forecasting integrated circuit output using multivariate grey model and grey relational analysis

This work proposes a new prediction approach using the multivariate grey model combined with grey relational analysis and results indicate that the proposed a combined method provides significantly improvement over the traditional methods in both forecasting techniques and in scanning variables.

Abstract

Integrated circuit outputs forecasts are of great economic value for both public and the private sectors in the world economy. In order to adjust policies and correct the invest portfolios; information relating to the development of the integrated circuit industry is of great importance to the governments, the policymakers, and stakeholders. Recently, numerous researchers have presented a wide range of forecasting techniques for the high-technology industry. Most of these methods are based on statistical studies that involve the use of univariate time series data. However, these models usually fail to clearly explain the insights of forecast results. This work proposes a new prediction approach using the multivariate grey model combined with grey relational analysis. Results indicate that the proposed a combined method provides significantly improvement over the traditional methods in both forecasting techniques and in scanning variables. Managerial implications and future research directions are also discussed.