A note on the statistical properties of nonparametric trend estimators by means of smoothing matrices
Estela Dagum and
Alessandra Luati
Journal of Nonparametric Statistics, 2009, vol. 21, issue 2, 193-205
Abstract:
Smoothing matrices associated with linear filters for the estimation of time series’ unobserved components differ from those used in linear regression or generalised additive models due to asymmetry. In fact, while projection smoother matrices are in general symmetric, filtering matrices are not. It follows that many inferential properties developed for symmetric projection matrices no longer hold for time-series smoothing matrices. However, the latter have a well-defined algebraic structure that allows one to derive many properties useful for inference in smoothing problems.In this note, some properties of symmetric smoother matrices are extended to centrosymmetric smoothing matrices. A decomposition of smoothing matrices in submatrices associated with the symmetric and asymmetric components of a filter enables us to consider the different assumptions that characterise estimation in the interior and at the boundaries of a finite time series.Matrix-based measures are defined to approximate the bias and the variance of a trend estimator, both in the interior and at the boundaries. These measures do not depend on the data and provide useful information on the bias–variance trade-off that affects non-parametric estimators.
Date: 2009
References: View complete reference list from CitEc
Citations:
Downloads: (external link)
http://hdl.handle.net/10.1080/10485250802534986 (text/html)
Access to full text is restricted to subscribers.
Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.
Export reference: BibTeX
RIS (EndNote, ProCite, RefMan)
HTML/Text
Persistent link: https://EconPapers.repec.org/RePEc:taf:gnstxx:v:21:y:2009:i:2:p:193-205
Ordering information: This journal article can be ordered from
http://www.tandfonline.com/pricing/journal/GNST20
DOI: 10.1080/10485250802534986
Access Statistics for this article
Journal of Nonparametric Statistics is currently edited by Jun Shao
More articles in Journal of Nonparametric Statistics from Taylor & Francis Journals
Bibliographic data for series maintained by Chris Longhurst ().