多因素时间序列的灰色预测模型  

Research on grey forecasting model of multi-factor time series

作　　者：苏变萍[1] 曹艳平[1] 王婷[1] 

机构地区：[1]西安建筑科技大学理学院,陕西西安710055

出　　处：《西安建筑科技大学学报（自然科学版）》2007年第2期289-292,共4页Journal of Xi'an University of Architecture & Technology(Natural Science Edition)

基　　金：陕西省教育厅专项基金项目(01JK133)

摘　　要：对于传统的单因素时间序列预测法在实际应用中的不足之处,提出采用灰色DGM(1,1)模型和多元线性回归原理相结合的方法,综合各种因素建立多因素时间序列的灰色预测模型。它首先利用DGM(1,1)模型对影响事物发展趋势的各项因素进行预测;然后利用多元线性回归法将各种因素综合起来,以预测事物的发展趋势。最后将该模型应用于预测分析陕西省的就业状况,取得了较好的预测效果,同时也验证了此模型的可行性。Considering the deficieney of the pracrical use of traditional forecasting me thod of single-factor time series, a grey forecasting method model for a mudti-factor time series is established by using DGM（1,1）model hased on the principle of multi-element linear regression to sgnthesige the various factors. The paper firstly forecasts the factors that affect the developing trend by using DGM（1,1）model,and further predicts the general development trend by using the method of multi-element linear regression and synthesizing the various factors. Finally, by forecasting and analysing the employment in Shaanxi province,the model is checked,obfaining some better results.

关 键 词：时间序列 单因素 多因素 预测模型 

分 类 号：TB114[理学—概率论与数理统计]