We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:


Current browse context:


Change to browse by:

References & Citations


(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Statistics > Methodology

Title: The De-Biased Whittle Likelihood

Abstract: The Whittle likelihood is a widely used and computationally efficient pseudo-likelihood. However, it is known to produce biased parameter estimates for large classes of models. We propose a method for de-biasing Whittle estimates for second-order stationary stochastic processes. The de-biased Whittle likelihood can be computed in the same $\mathcal{O}(n\log n)$ operations as the standard approach. We demonstrate the superior performance of the method in simulation studies and in application to a large-scale oceanographic dataset, where in both cases the de-biased approach reduces bias by up to two orders of magnitude, achieving estimates that are close to exact maximum likelihood, at a fraction of the computational cost. We prove that the method yields estimates that are consistent at an optimal convergence rate of $n^{-1/2}$, under weaker assumptions than standard theory, where we do not require that the power spectral density is continuous in frequency. We describe how the method can be easily combined with standard methods of bias reduction, such as tapering and differencing, to further reduce bias in parameter estimates.
Comments: To appear shortly in Biometrika. Full published version includes extensions of theory to non-Gaussian processes, and new simulation examples with an AR(4) and non-Gaussian process
Subjects: Methodology (stat.ME); Statistics Theory (math.ST); Computation (stat.CO); Machine Learning (stat.ML)
Cite as: arXiv:1605.06718 [stat.ME]
  (or arXiv:1605.06718v3 [stat.ME] for this version)

Submission history

From: Adam Sykulski Dr [view email]
[v1] Sun, 22 May 2016 00:47:52 GMT (88kb)
[v2] Wed, 22 Nov 2017 23:38:15 GMT (202kb)
[v3] Wed, 12 Sep 2018 15:39:43 GMT (202kb)

Link back to: arXiv, form interface, contact.