2001
DOI: 10.1198/016214501753168118
|View full text |Cite
|
Sign up to set email alerts
|

Bayesian Wavelet Regression on Curves With Application to a Spectroscopic Calibration Problem

Abstract: Motivated by calibration problems in near-infrared (N IR) spectroscopy, we consider the linear regression setting in which the many predictor variables arise from sampling an essentially continuous curve at equally spaced points and there may be multiple predictands. We tackle this regression problem by calculating the wavelet transforms of the discretized curves, then applying a Bayesian variable selection method using mixture priors to the multivariate regression of predictands on wavelet coef cients. For pr… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
4
1

Citation Types

2
119
0
4

Year Published

2007
2007
2015
2015

Publication Types

Select...
8
1

Relationship

0
9

Authors

Journals

citations
Cited by 157 publications
(125 citation statements)
references
References 28 publications
2
119
0
4
Order By: Relevance
“…We consider the biscuit dough dataset, which was previously analyzed by Brown et al [1] and more recently by [11]. The dataset was obtained from a near-infrared (NIR) spectroscopy experiment used to analyze the composition of biscuit dough pieces, and it is available as part of the R package ppls [17].…”
Section: Biscuit Dough Datamentioning
confidence: 99%
See 1 more Smart Citation
“…We consider the biscuit dough dataset, which was previously analyzed by Brown et al [1] and more recently by [11]. The dataset was obtained from a near-infrared (NIR) spectroscopy experiment used to analyze the composition of biscuit dough pieces, and it is available as part of the R package ppls [17].…”
Section: Biscuit Dough Datamentioning
confidence: 99%
“…Email addresses: joyee-ghosh@uiowa.edu (Joyee Ghosh), aixin-tan@uiowa.edu (Aixin Tan) 1 The authors contributed equally. 2 Supplementary material includes details of algorithms, statement and proof of a lemma, and theoretical comparisons of sandwich algorithms in some examples.…”
Section: Introductionmentioning
confidence: 99%
“…The advantage of introducing Bayesian methodology into neural networks was reviewed [10] through the application to the calibration of near infrared spectroscopy. As early advocates of Bayesian approach, Brown and co-workers proposed a number of calibration models that demonstrated promising results, such as Bayesian variable selection methods for calibration [21,22] and wavelet regression [23]. One salient feature of Brown's work is that Markov chain Monte Carlo (MCMC) simulation is employed for the inference of the model parameters.…”
Section: Introductionmentioning
confidence: 99%
“…Models are then built by applying regression techniques to the wavelet coefficients resulting from the bank filter decomposition [4,6,23]. In many applications, such a procedure led to improvements in the prediction performance of the resulting models [24].…”
Section: Introdutionmentioning
confidence: 99%