PDF(806 KB)
PDF(806 KB)
PDF(806 KB)
波罗的海干散货指数预测的非等间隔灰色波形预测方法
Forecasting Baltic dry index with unequal-interval grey wave forecasting model
为提高对波罗的海干散货指数(BDI)预测的准确度,根据BDI指数这类波动幅度不规律的时间序列的特征,提出一种非等间隔的灰色波形预测方法,即应用分位数法选取非等间隔的等高线,并有选择地对等高时刻序列进行GM(1,1)建模.通过对BDI指数月数据的建模与预测表明,非等间隔的灰色波形预测方法较传统的灰色波形预测方法和ARMA(1,1)模型在预测精度和运算效率方面具有明显优势.
In order to improve the prediction accuracy of time series with irregular fluctuation ranges, such as Baltic Dry Index, this paper proposes a unequal-interval grey wave forecasting model, particularly, using quantile to define the contour lines and establishing GM(1,1) models only based on qualified isoheight time sequences. In the end, this paper uses Baltic Dry Index (BDI) to testify the modified model empirically. The empirical analysis indicates the unequal-interval grey wave forecasting model is better than traditional grey wave forecasting model in prediction accuracy and computational efficiency.
波罗的海干散货指数 / 波动幅度不规律 / 非等间隔 / 灰色波形预测
unequal-Interval / greywave forecasting / Baltic Dry Index / irregular fluctuations range
/
| 〈 |
|
〉 |