The development of an interval grey regression model for limited time series forecasting

Abstract

The grey model GM(1, 1) is a popular forecasting method in management and engineering applications. In order to obtain more validity and reliable forecasting values, fuzzy grey regression model is proposed by hybridizing fuzzy set into grey model GM(1, 1) for solving crisp-input and fuzzy-input limited time series. In the present study, we continue to investigate and derive the interval grey regression model with a minimum regression interval by necessity analysis in which the predicted interval is possibly to be belonged to the collected interval data with certain membership degree. In the illustrative example we did not have as much information as the research institutes of Display search, iSuppli, and PIPA, but we succeeded at forecasting LCD TV demand by using the interval grey regression model with limited and interval-input data.