Optimized Grey Derivative of GM (1, 1)

Abstract

From the production of GM (1,1) grey derivative, this article arguments logically the rationality of using weighted average of forward difference quotient and backward difference quotient as GM(1,1) grey derivative whitenization value in the theories. It gives the concrete expression type of weighted coefficient and builds up a new GM(1,1) model. It proves that the new model has the white exponential coincidence law in theory and puts forward a new method to solve parameters of the new model. Simulation and prediction of practice examples show that this model and method are useful and effective.