摘要
将组合预测法应用于我国GDP的预测,以提高预测精度。通过赋予合理权重,将指数平滑模型、拟合模型、ARIMA模型和支持向量回归模型加权组合。对各模型进行平均绝对百分误差(MAPE)、均方根误差(RESE)和希尔不等系数(Theil IC)等指标的比较,证明单一模型经过组合能够提高预测精度。
Combination forecasting model will be applied to forecast the gross domestic product of China,in order to make the result more exactly.The model which combined exponential smoothing model,fitting model,ARIMA model and SVR model can improve the accuracy of fix and forecast by proper weighs.By analyzing different indicators,the author consider that combination forecasting model is superior to four other models in this application.
出处
《廊坊师范学院学报（自然科学版）》
2010年第2期87-89,共3页
Journal of Langfang Normal University(Natural Science Edition)
基金
辽宁省高等学校科研项目资助(2008343)
