Improving autocorrelation regression for the Hurst parameter estimation of long-range dependent time series based on golden section search

This article presents an improved autocorrelation correlation function (ACF) regression method of estimating the Hurst parameter of a time series with long-range dependence (LRD) by using golden section search (GSS). We shall show that the present method is substantially efficient than the conventional ACF regression method of H estimation. Our research uses fractional Gaussian noise as a data case but the method introduced is applicable to time series with LRD in general.