Posterior contraction and credible sets for filaments of regression functions

DOI10.1214/20-EJS1705MaRDI QIDQ2180076zbMATHFDO

Authors Yanyan Li

Publication date 13 May 2020

Published in Electronic Journal of Statistics (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1803.03898

zbMATH Keywords

nonparametric regressionB-splinesposterior contractionfilamentcoveragecredibility

Mathematics Subject Classification ID

Nonparametric regression and quantile regression (62G08) Asymptotic properties of nonparametric inference (62G20) Nonparametric tolerance and confidence regions (62G15) Numerical computation using splines (65D07) Applications of statistics to physics (62P35) Seismology (including tsunami modeling), earthquakes (86A15)

Abstract: A filament consists of local maximizers of a smooth function $f$ when moving in a certain direction. A filamentary structure is an important feature of the shape of an object and is also considered as an important lower dimensional characterization of multivariate data. There have been some recent theoretical studies of filaments in the nonparametric kernel density estimation context. This paper supplements the current literature in two ways. First, we provide a Bayesian approach to the filament estimation in regression context and study the posterior contraction rates using a finite random series of B-splines basis. Compared with the kernel-estimation method, this has a theoretical advantage as the bias can be better controlled when the function is smoother, which allows obtaining better rates. Assuming that $f:mathbb{R}^{2}mapstomathbbR$ belongs to an isotropic H"{o}lder class of order $alphageq4$, with the optimal choice of smoothing parameters, the posterior contraction rates for the filament points on some appropriately defined integral curves and for the Hausdorff distance of the filament are both $(n/logn{)}^{(2-alpha)/(2(1+alpha))}$. Secondly, we provide a way to construct a credible set with sufficient frequentist coverage for the filaments. We demonstrate the success of our proposed method in simulations and one application to earthquake data.

Recommendations

• Theoretical analysis of nonparametric filament estimation
• The geometry of nonparametric filament estimation
• On the path density of a gradient field
• Posterior contraction and credible regions for level sets
• Supremum norm posterior contraction and credible sets for nonparametric multivariate regression

Cites work

• scientific article; zbMATH DE number 5190601 (Why is no real title available?)
• A practical guide to splines
• Adaptive Bernstein-von Mises theorems in Gaussian white noise
• Adaptive multiscale detection of filamentary structures in a background of uniform random points
• Anti-concentration and honest, adaptive confidence bands
• Asymptotic normality of plug-in level set estimates
• Asymptotic theory for density ridges
• Bayesian inverse problems with Gaussian priors
• Bayesian mode and maximum estimation and accelerated rates of contraction
• Bootstrap confidence bands for regression curves and their derivatives
• Confidence bands in density estimation
• Confidence regions for level sets
• Confidence regions for means of random sets using oriented distance functions
• Empirical Bayes oracle uncertainty quantification for regression
• Estimation and Confidence Regions for Parameter Sets in Econometric Models
• Frequentist coverage of adaptive nonparametric Bayesian credible sets
• Gaussian approximations and multiplier bootstrap for maxima of sums of high-dimensional random vectors
• Gronwall, Bellman and Pachpatte type integral inequalities with applications
• Inference on sets in finance
• Integral curves of noisy vector fields and statistical problems in diffusion tensor imaging: nonparametric kernel estimation and hypotheses testing
• Least squares estimators of the mode of a unimodal regression function
• Locally defined principal curves and surfaces
• Manifold estimation and singular deconvolution under Hausdorff loss
• Needles and straw in a haystack: robust confidence for possibly sparse sequences
• Nonparametric estimation of the location of a maximum in a response surface.
• Nonparametric ridge estimation
• On Bayesian supremum norm contraction rates
• On coverage and local radial rates of credible sets
• Rates of contraction for posterior distributions in \(L^{r}\)-metrics, \(1 \leq r \leq \infty\)
• Ridges in image and data analysis
• Supremum norm posterior contraction and credible sets for nonparametric multivariate regression
• Theoretical analysis of nonparametric filament estimation
• Uncertainty quantification for the horseshoe (with discussion)
• Weak convergence and empirical processes. With applications to statistics

Cited in

(7)

• scientific article; zbMATH DE number 788279 (Why is no real title available?)
• The geometry of nonparametric filament estimation
• On the path density of a gradient field
• Bayesian consistency with the supremum metric
• Posterior contraction and credible regions for level sets
• Asymptotic confidence regions for density ridges
• Theoretical analysis of nonparametric filament estimation

This page was built for publication: Posterior contraction and credible sets for filaments of regression functions

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2180076)