摘要
提出了基于自回归 (AR)模型对时间序列统一建模的新观点和方法 ,可大大减少计算量 ,并在微机上编程实现 .以实例对动态模型与静态模型分别作了应用比较 ,结果表明 ,时间序列分析动态模型是系统分析的重要方法 ,是统计预测法中的高级预测方法 ,预测精度高 ,用途广泛 .而静态模型适合于内插 ,不适合于外推预报 .
This article issues a new viewpoint and method in modeling for time series based on AR model. The method is able to give less calculating and to be programmed on computer. Compared with the cases of application between the dynamic model and the static model, it is indicated that the model of time series analysis dynamic model is an important and advanced forecasting method in statistical forecasting, and has higher forecasting accuracy. This method can be widely used in fields. But static model is available for interpolation, not for extrapolated forecasting.
出处
《同济大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2001年第3期294-298,共5页
Journal of Tongji University:Natural Science
