On property of derived grey Verhulst model under multiple transformation of modelling data

Abstract

Multiple transformation is one of the essential means of reducing the complexity of grey modeling, aiming to reveal the changing law of modeling accuracy of the derived grey Verhulst model when multiplication transformations acting on its modeling sequence and enhance its modeling performance. This paper discusses the parameter characteristics of the grey derived Verhulst model under multiple transformations, and demonstrates its effect on its simulative value and predictive value by investigating the multiple transformations acting on the raw data sequence of this grey model. The research findings indicate that the modeling accuracy of the derived grey Verhulst model has no relationship with multiple transformations of raw data sequences of systems. The research conclusion implies that the data level can be reduced and the course of building models can be simplified, but simulative accuracy and predictive accuracy of this model remain unchanged.