Application of wavelet analysis in the estimation of Hurst index
Jian Hou · Jounal of Xidian University · 2002
The concept of time verying Hurst index H(t) is put forward in the biased stochastic process. Wavelet has an advantage of analying the locally self similar process. Based on it, we construct our algorithm. The estimation formula (t) of Hurst index is proposed for treating sample data through Daubechies wavelet, proving that (t) is a consistent result of H(t).