Research of modeling of multi-variable non-equidistant MGM(1,m) model
Research of modeling of multi-variable non-equidistant MGM(1,m) model
复制标题
DOI：
--
复制
发表时间：
2011
期刊：
影响因子：
--
通讯作者：
Xiong Ping-ping;Dang Yao-guo;Zhu Hui
中科院分区：
文献类型：
--
作者：
Xiong Ping-ping;Dang Yao-guo;Zhu Hui
针对多变量相互作用的非等距原始数据序列的模拟预测问题，建立了非等距情况下的多变量MGM（1，λ）模型，构造了非等距MGM（1，λ）模型，以提高模拟预测精度。应用实例表明，多变量非等距MGM（1，1）模型比单变量非等距GM（1，1）模型具有更高的模拟和预测精度。该模型适用于非等距建模和等距建模，因此非等距模型具有广泛的应用价值。
Based on the problem of simulation and prediction of the non-equidistant raw data series for the multi-variable with interaction, the multi-variable MGM(1,�� ) model in the case of non-equidistance is modeled and the non-equidistance MGM(1,�� ) model is constructed in order to improve the accuracy of simulation and prediction. The simulation and prediction accuracy is proved to be higher in multi-variable non-equidistant MGM(1,�� ) model than in single-variable non- equidistant GM(1,1) models according to an application. The proposed model is suitable in non-equidistant modeling and equidistant modeling, thus, non-equidistant model has broad applying value.