Parameter Optimization Methods for Grey Prediction Models

The parameters of grey prediction models are important factors that affect their simulation and prediction performance. Grey prediction models can be divided into two categories based on the functions of parameters: process parameters (initial value, background value, accumulative order) and basic parameters (development coefficient, grey action quantity, etc.). Process parameters determine the size of basic parameters by Matrix B and Matrix Y. Basic parameters are important parameters that affect the performance of grey prediction models, and are generally estimated through the least square method. The specific process has been introduced in the previous chapters. In this chapter, some optimization methods for process parameters of grey prediction models will be introduced, aiming at building a grey prediction model with high-precision and better performance.