Bayesian registration of functions and curves

Bayesian analysis of functions and curves is considered, where warping and other geometrical transformations are often required for meaningful comparisons. The functions and curves of interest are represented using the recently introduced square root velocity function, which enables a warping invar...

Full description

Bibliographic Details
Main Authors: Cheng, Wen, Dryden, Ian L., Huang, Xianzheng
Format: Article
Published: International Society for Bayesian Analysis 2015
Subjects:
Online Access:https://eprints.nottingham.ac.uk/29193/
_version_ 1848793733788598272
author Cheng, Wen
Dryden, Ian L.
Huang, Xianzheng
author_facet Cheng, Wen
Dryden, Ian L.
Huang, Xianzheng
author_sort Cheng, Wen
building Nottingham Research Data Repository
collection Online Access
description Bayesian analysis of functions and curves is considered, where warping and other geometrical transformations are often required for meaningful comparisons. The functions and curves of interest are represented using the recently introduced square root velocity function, which enables a warping invariant elastic distance to be calculated in a straightforward manner. We distinguish between various spaces of interest: the original space, the ambient space after standardizing, and the quotient space after removing a group of transformations. Using Gaussian process models in the ambient space and Dirichlet priors for the warping functions, we explore Bayesian inference for curves and functions. Markov chain Monte Carlo algorithms are introduced for simulating from the posterior. We also compare ambient and quotient space estimators for mean shape, and explain their frequent similarity in many practical problems using a Laplace approximation. Simulation studies are carried out, as well as practical alignment of growth rate functions and shape classification of mouse vertebra outlines in evolutionary biology. We also compare the performance of our Bayesian method with some alternative approaches.
first_indexed 2025-11-14T19:05:00Z
format Article
id nottingham-29193
institution University of Nottingham Malaysia Campus
institution_category Local University
last_indexed 2025-11-14T19:05:00Z
publishDate 2015
publisher International Society for Bayesian Analysis
recordtype eprints
repository_type Digital Repository
spelling nottingham-291932020-05-04T20:10:43Z https://eprints.nottingham.ac.uk/29193/ Bayesian registration of functions and curves Cheng, Wen Dryden, Ian L. Huang, Xianzheng Bayesian analysis of functions and curves is considered, where warping and other geometrical transformations are often required for meaningful comparisons. The functions and curves of interest are represented using the recently introduced square root velocity function, which enables a warping invariant elastic distance to be calculated in a straightforward manner. We distinguish between various spaces of interest: the original space, the ambient space after standardizing, and the quotient space after removing a group of transformations. Using Gaussian process models in the ambient space and Dirichlet priors for the warping functions, we explore Bayesian inference for curves and functions. Markov chain Monte Carlo algorithms are introduced for simulating from the posterior. We also compare ambient and quotient space estimators for mean shape, and explain their frequent similarity in many practical problems using a Laplace approximation. Simulation studies are carried out, as well as practical alignment of growth rate functions and shape classification of mouse vertebra outlines in evolutionary biology. We also compare the performance of our Bayesian method with some alternative approaches. International Society for Bayesian Analysis 2015 Article PeerReviewed Cheng, Wen, Dryden, Ian L. and Huang, Xianzheng (2015) Bayesian registration of functions and curves. Bayesian Analysis, 2015 . ISSN 1936-0975 Ambient space Dirichlet Gaussian process Quotient space Shape Warp. http://projecteuclid.org/euclid.ba/1433162661 doi:10.1214/15-BA957 doi:10.1214/15-BA957
spellingShingle Ambient space
Dirichlet
Gaussian process
Quotient space
Shape
Warp.
Cheng, Wen
Dryden, Ian L.
Huang, Xianzheng
Bayesian registration of functions and curves
title Bayesian registration of functions and curves
title_full Bayesian registration of functions and curves
title_fullStr Bayesian registration of functions and curves
title_full_unstemmed Bayesian registration of functions and curves
title_short Bayesian registration of functions and curves
title_sort bayesian registration of functions and curves
topic Ambient space
Dirichlet
Gaussian process
Quotient space
Shape
Warp.
url https://eprints.nottingham.ac.uk/29193/
https://eprints.nottingham.ac.uk/29193/
https://eprints.nottingham.ac.uk/29193/