1. Institute of Grey System Studies, Nanjing University of Aeronautics and Astronautics, 29, Yudao St., Nanjing, Jiangsu Province 210016, PR China;2. Centre for Computational Intelligence, De Montfort University, Leicester LEl 9BH, UK
Abstract:
To solve the problem that traditional grey models were constructed on the hypothesis that the original data sequence is in accord with homogeneous index trend rather than non-homogeneous index trend. A novel grey forecasting model based on non-homogeneous index sequence approximately (abbreviated as NDGM) is proposed. It is proved that the models based on homogeneous index sequence are all special cases that of non-homogeneous index sequence. The recursive function of the NDGM model is proposed and the forecasting precision of the model based on pure non-homogeneous index sequence is discussed and the affine properties of the model are further studied. Finally, one numerical case is used to show the effective results of the proposed model.