A Bayesian wavelet estimation method for estimating parameters of a stationary I(d) process is represented as an useful alternative to the existing frequentist wavelet estimation methods. The effectiveness of the proposed method is demonstrated through Monte Carlo simulations. The sampling from the posterior distribution is through the Markov Chain Monte Carlo (MCMC) easily implemented in the WinBUGS software package.
Material in this article is copyrighted and initially published by the Journal of Modern Applied Statistical Methods, 2005, Volume 4, Issue 1, 140-154, ISSN 1538-9472, Wayne State University, College of Education, 5425 Gullen Mall, Detroit, MI, 48202. http://www.coe.wayne.edu. http://education.wayne.edu/jmasm/
Qu, Leming. (2005). "Bayesian Wavelet Estimation of Long Memory Parameter". Journal of Modern Applied Statistical Methods, 4(1), 140-154.