Estimation of the Hurst exponent from noisy data: a Bayesian approach
Interdisciplinary Center for Dynamics of Complex Systems, University of Potsdam, Karl-Liebknecht-Str. 24, 14476 Potsdam, Germany
Received: 13 March 2012
Received in final form: 16 May 2012
Published online: 6 August 2012
We consider a model based on the fractional Brownian motion under the influence of noise. We implement the Bayesian approach to estimate the Hurst exponent of the model. The robustness of the method to the noise intensity is tested using artificial data from fractional Brownian motion. We show that estimation of the parameters achieved when noise is considered explicitly in the model. Moreover, we identify the corresponding noise-amplitude level that allow to receive the correct estimation of the Hurst exponents in various cases.
Key words: Statistical and Nonlinear Physics
© EDP Sciences, Società Italiana di Fisica and Springer-Verlag, 2012