科研之友
微信
新浪微博
Facebook
分享链接
摘要
This paper presents a novel multi-variable grey prediction model GMI(1,N) based on grey incidence analysis. In the proposed forecasting model relationships between variables ignored in the current literature, is considered Grey incidence degree (GID) is used to express the relationships between variables, and for each relevant sequence, there is a GID sequence to describe its relationship with the feature sequence. The multi-variable grey prediction model GMI(1,N) is established by considering added GID sequences. To check the effectiveness and feasibility, the case of Sino-Russian Timber Trade Volume Forecasting is presented Several conclusions are shown in the case. First, AME decreases with the increasing size of relevant factors until AME=0 and for both GM(1,N) and GMI(1,N) models, AME=0 is a mutational point. Second, before AME=0, the accuracy of GMI(1,N) is higher than that of GM(1,N). Third, from the viewpoint of AME, the GMI(1,N) model converges better than the GM(1,N) model.
