基于奇异谱分析的上证指数预测模型

Singular Spectrum Analysis and Forecast Model Based on Shanghai Composite Index

该文应用奇异谱分析(SSA)的方法,对一维时间序列——证券指数,构造延迟矩阵,运用时间经验正交函数(EOF),对原序列进行重构,有效地提取序列中隐含的波形信号;利用自回归移动平均模型(ARMA模型)对原序列和重构结果分别进行预测,并进行比较。

According to a single variate time series about SCI, this paper build a delay matrix and restructure a new single variate time series applying SSA. Puck up the distinct information of the old time series. Then give the forecast index of the new time series, and compare with the old. The result is satisfactory.