Wavelet-Based Exact Maximum Likelihood Estimation of Arfima (p,d,q) Parameters

In this paper, we modify wavelets so that wavelet coefficients become statistically independent. Under standard regularity conditions, we determine the distribution of the wavelet coefficients. Using the moving average representation of the ARFIMA(p,d,q) process, we obtain wavelet-based exact maximum likelihood estimators of the ARFIMA(p,d,q) parameters. The corresponding Fisher information matrix is also derived

MoreLess

Year of publication:	[2009]
Authors:	Gonzaga, Alex
Other Persons:	Bonzo, Daniel (contributor)
Publisher:	[2009]: [S.l.] : SSRN
Subject:	Maximum-Likelihood-Schätzung \| Maximum likelihood estimation \| Schätztheorie \| Estimation theory \| ARMA-Modell \| ARMA model
Description of contents:	Abstract [papers.ssrn.com]