Smoothing locally regular processes by Bayesian nonparametric methods, with applications to acid rain data analysis

Smoothing locally regular processes by Bayesian nonparametric methods, with applications to acid rain data analysis

We consider the problem of recovering an unknown smooth function from the data using the Bayesian nonparametric approach proposed by Weerahandi and Zidek (1985). Selected nonparametric smoothing methods are reviewed and compared with this new method. At each value of the independent variable, the sm...

Full description

Bibliographic Details
Main Author: Ma, Hon Wai
Language:English
Published: University of British Columbia 2010
Online Access:http://hdl.handle.net/2429/26004
id ndltd-UBC-oai-circle.library.ubc.ca-2429-26004
record_format oai_dc
spelling ndltd-UBC-oai-circle.library.ubc.ca-2429-260042018-01-05T17:43:26Z Smoothing locally regular processes by Bayesian nonparametric methods, with applications to acid rain data analysis Ma, Hon Wai We consider the problem of recovering an unknown smooth function from the data using the Bayesian nonparametric approach proposed by Weerahandi and Zidek (1985). Selected nonparametric smoothing methods are reviewed and compared with this new method. At each value of the independent variable, the smooth function is assumed to be expandable in a Taylor series to the pth order. Two methods, cross-validation and "backfitting" are used to estimate the a priori unspecified hyperparameters. Moreover, a data-based procedure is introduced to select the appropriate order p. Finally, an analysis of an acid-rain, wet-deposition time series is included to indicate the efficacy of the proposed methods. Science, Faculty of Statistics, Department of Graduate 2010-06-27T16:53:36Z 2010-06-27T16:53:36Z 1986 Text Thesis/Dissertation http://hdl.handle.net/2429/26004 eng For non-commercial purposes only, such as research, private study and education. Additional conditions apply, see Terms of Use https://open.library.ubc.ca/terms_of_use. University of British Columbia
collection NDLTD
language English
sources NDLTD
description We consider the problem of recovering an unknown smooth function from the data using the Bayesian nonparametric approach proposed by Weerahandi and Zidek (1985). Selected nonparametric smoothing methods are reviewed and compared with this new method. At each value of the independent variable, the smooth function is assumed to be expandable in a Taylor series to the pth order. Two methods, cross-validation and "backfitting" are used to estimate the a priori unspecified hyperparameters. Moreover, a data-based procedure is introduced to select the appropriate order p. Finally, an analysis of an acid-rain, wet-deposition time series is included to indicate the efficacy of the proposed methods. === Science, Faculty of === Statistics, Department of === Graduate
author Ma, Hon Wai
spellingShingle Ma, Hon Wai
Smoothing locally regular processes by Bayesian nonparametric methods, with applications to acid rain data analysis
author_facet Ma, Hon Wai
author_sort Ma, Hon Wai
title Smoothing locally regular processes by Bayesian nonparametric methods, with applications to acid rain data analysis
title_short Smoothing locally regular processes by Bayesian nonparametric methods, with applications to acid rain data analysis
title_full Smoothing locally regular processes by Bayesian nonparametric methods, with applications to acid rain data analysis
title_fullStr Smoothing locally regular processes by Bayesian nonparametric methods, with applications to acid rain data analysis
title_full_unstemmed Smoothing locally regular processes by Bayesian nonparametric methods, with applications to acid rain data analysis
title_sort smoothing locally regular processes by bayesian nonparametric methods, with applications to acid rain data analysis
publisher University of British Columbia
publishDate 2010
url http://hdl.handle.net/2429/26004
work_keys_str_mv AT mahonwai smoothinglocallyregularprocessesbybayesiannonparametricmethodswithapplicationstoacidraindataanalysis
_version_ 1718592965980454912

Similar Items