Abstract: To analyze the factor of time which affects the combined forecasting's prediction accuracy, based on all of the forecasting method's prediction accuracy stability, reference for the modeling thought of one—dimension AR(p)model, we present the combined forecasting based on the factor of time. The example verification shows that the prediction accuracy of the combined forecasting based on the factor of time is higher than the single forecast method, and if the time span of samples is longer, the prediction accuracy is higher.

Key words: combined forecasting, factor of time, AR(p)model, precision

摘要： 为了分析时间因素对组合预测模型预测精度的影响,基于对长期内各预测方法预测精度稳定性的考虑,借鉴一维AR(p)模型建模思路,提出了基于时间因素的组合预测模型建模方法。实例验证表明:相对于单个预测方法,考虑时间因素的组合预测方法的预测精度更高,且样本时间跨度越长,预测精度也越高。

关键词: 组合预测, 时间要素, AR(p)模型, 精度