多重分形谱主要系数在股票预测问题中的研究

The Research of Multifractal Spectrum's Main Parameters in Stocks' Forecast

本文根据分形理论及多重分形原理,利用解析的方法对离散数据的多重分形谱的主要参数—奇异标度指数α及奇异谱函数f(α)进行了计算;并把该方法用于研究上证大盘A股综合指数以及其他多只个股股价的5分钟高频时间序列,根据大量实验结果分析,提出了可以依据股票股价最近时间段的多重分形谱主要参数变化情况来预测股票股价短期走势的初步结论,为股票预测分析提供了一定的参考依据。

Based on the fractal theory and multi-fractaltheory, this paper uses analytical methods to calculatethe main parameters αanomalousscaling exponent andsingular spectrum function f（α）of thediscretedata multifractal spectrum, and uses this method to study evidence tape Ashares comprehensive index and 5 minutes high frequency time series of otherstock price.Accordingto a large number of experimental results, it is put forward that the short-termmovements of stock price can be predicted by the multi-fractal spectrum of themain parameters of the stock price changes in recent period, so as to provide thereference for the stock prediction.