基于SSA的金融时间序列自适应分解预测

提出了分解预测的思想,通过SSA将序列分解成低频与高频两部分,分别采用最小均方(LMS)自适应自回归移动平均(ARIMA)与LMS自适应自回归(AR)模型进行预测,然后将两者叠加便可得原始序列预测值.同时,为了更好地捕捉序列局部突变,缩减预测延迟,提高预测精度,对EaLMS算法(基于误差调整的LMS算法)参数进行修正并...

Financial Time Series Adaptive Decomposition Prediction Based on SSA

This paper proposed a decomposition method, in which the sequence is divided into two parts：the low and the high frequency sequence. Then the two parts were predicted by LMS adaptive ARIMA and LMS adaptive AR model respective- ly, and the final result was obtained by composing the two predictors. Meanwhile, in order to capture the sequence partial mu- tation well, reduce the prediction delay, and improve the prediction precision, the modified EaLMS algorithm was applied in the decomposition prediction. The experimental result shows that, compared with direct LMS prediction, the modified decomposition method has better prediction performance.