子波变换及其在股市数据分析中的应用

把股票日收益率看作一维时间信号，通过子波变换与多尺度分析，求出了体现股市数据奇性的Lipschitz指数α。α为负表明在奇点比不连续更加奇异，证明了股票价格的变化是分形的。同时，由文中图示可知，在尺度s很大时，子波变换和多尺度分析可以将股市数据中偶然因素造成的涨跌消除，具有突出主要因素和客观突变点的特点，这一点对于从宏观上预测股价的走势有重要意义。

Wavelet Transform and Its Application in Stock Market Data Analysis

A lipschitz exponent α, which refle cts the singlarity nature ofstock data, is obtained with wavelet trans form and muti-scale analysis by regarding stock day profit ratio as a one -dimentional time signal. Α in this paper is negative, showing that the singul arity is more singular than non-continuum. This proves that the variation of st ock prices is fractal. At the sametime, it is pointed out that when scales is very big, wavelet transform and multi-scale analysis can remove the up-an d-down of stock market data caused by accidental factors and give prominence to primary factors and macroscopic sudden change points. This is important in predicting the variat ion trend of stock prices from macroscopic aspect.