摘要
本文用我国城镇居民 1992— 1998年消费数据分别建立各年度的扩展线性支出系统模型 ,用模型预测值与实际发生值进行对比 ,从而检验扩展线性支出系统的预测精度。检验结果表明 ,预测误差随被预测年的推移而迅速增大。并对误差产生的原因作了分析。最后得出结论 :扩展线性支出系统一般只宜用作消费需求的短期预测。
This paper first build ELES models respectively using Chinese consumption disbursal data of denizen of towns from 1982 to 1998,and then forecast the future values of consumption disbursal with these models,finally compare forecast values with sooth values to exploring the forecasting precision of ELES.The result shows that the error fleetly become more notable as forecasting year increasing.Afterward the paper theoretically analyzed the causation of error producing.Finally educe conclusion:the ELES currently can only be used for short time consumption forecast.
出处
《数理统计与管理》
CSSCI
北大核心
2002年第5期37-41,共5页
Journal of Applied Statistics and Management
