Analysis of stochastic time series in the presence of strong measurement noise
Abstract
An alternative approach for the analysis of Langevin-type stochastic processes in the presence of strong measurement noise is presented. For the case of Gaussian distributed, exponentially correlated, measurement noise it is possible to extract the strength and the correlation time of the noise as well as polynomial approximations of the drift and diffusion functions of the underlying Langevin equation.
- Publication:
-
Physical Review E
- Pub Date:
- February 2011
- DOI:
- 10.1103/PhysRevE.83.021113
- arXiv:
- arXiv:1010.5641
- Bibcode:
- 2011PhRvE..83b1113L
- Keywords:
-
- 05.40.Ca;
- 02.50.Ey;
- Noise;
- Stochastic processes;
- Physics - Data Analysis;
- Statistics and Probability
- E-Print:
- 12 pages, 10 figures