基于小波与动态GM(1,1)-ARIMA模型的股价预测研究

The Research of Stock Price Prediction Based on the Wavelet Dynamic GM(1,1)-ARIMA Model

针对传统时间序列股价预测模型的时间滞后性,提出一种基于小波与动态GM(1,1)-ARIMA的股价预测模型。运用小波分析对股价数据进行预处理,对小波重构序列建立ARIMA模型,考虑预测过程中未来因素对系统的影响,建立动态GM(1,1)模型。选取沪深300指数进行实证分析,结果表明所提出的小波与动态GM(1,1)-ARIMA模型与传统股价预测模型相比,其预测精度最高。

Aiming at th e t ime lag of th e t ra d i t io n a l t ime series s to ck price fo reprice forecast ing model based on wavelet and dynamic GM （1,1）-AR IMA model is p ro p osed . T h e datawere pret reated by wavelet analysis. Based on wavelet reconstruct ion sequences, th e A R IMA model was establ ished and the dynamic GM（ ？1） modeiconsidering the inf luence of future factors on the establ ished. We do the empirical analysis on Shanghai and shenzhen 300 rndex , th e re su l ts compared wi th the t radi t ional stock price forecast ing model, th e p ro p osed GM （ 1 , 1 ） -Athe highest prediction accuracy.