Bayesian estimation of self-similarity exponent
- In this study we propose a Bayesian approach to the estimation of the Hurst exponent in terms of linear mixed models. Even for unevenly sampled signals and signals with gaps, our method is applicable. We test our method by using artificial fractional Brownian motion of different length and compare it with the detrended fluctuation analysis technique. The estimation of the Hurst exponent of a Rosenblatt process is shown as an example of an H-self-similar process with non-Gaussian dimensional distribution. Additionally, we perform an analysis with real data, the Dow-Jones Industrial Average closing values, and analyze its temporal variation of the Hurst exponent.
Author details: | Natallia Makarava, Sabah Benmehdi, Matthias HolschneiderORCiDGND |
---|---|
DOI: | https://doi.org/10.1103/PhysRevE.84.021109 |
ISSN: | 1539-3755 |
Title of parent work (English): | Physical review : E, Statistical, nonlinear and soft matter physics |
Publisher: | American Physical Society |
Place of publishing: | College Park |
Publication type: | Article |
Language: | English |
Year of first publication: | 2011 |
Publication year: | 2011 |
Release date: | 2017/03/26 |
Volume: | 84 |
Issue: | 2 |
Number of pages: | 9 |
Funding institution: | Graduate School NADI, University of Potsdam |