An improved prediction model for interval grey number

In this paper, the parameter λ would be introduced to forecast grey number sequence that was based on the paper with known whitenization weight function published by Bo Zeng. We use variable weight instead of no-preference generation. So the most appropriate parameter to build the DGM (1,1) would be chose by GA. Then an improved grey prediction model for forecasting interval grey number is proposed. An application analysis is presented to illustrate the effectiveness and practicability of the proposed model. Simulation examples show the average relative error of the proposed model has been significantly improved.