An Empirical Bayes Approach to Shrinkage Estimation on the Manifold of Symmetric Positive-Definite Matrices

DOI10.1080/01621459.2022.2110877arXiv2007.02153OpenAlexW3038725864WikidataQ114101019 ScholiaQ114101019MaRDI QIDQ6153989FDOQ6153989

Authors: Chun-Hao Yang, Hani Doss, Baba C. Vemuri 

Publication date: 19 March 2024

Published in: Journal of the American Statistical Association (Search for Journal in Brave)

Abstract: The James-Stein estimator is an estimator of the multivariate normal mean and dominates the maximum likelihood estimator (MLE) under squared error loss. The original work inspired great interest in developing shrinkage estimators for a variety of problems. Nonetheless, research on shrinkage estimation for manifold-valued data is scarce. In this paper, we propose shrinkage estimators for the parameters of the Log-Normal distribution defined on the manifold of $NimesN$ symmetric positive-definite matrices. For this manifold, we choose the Log-Euclidean metric as its Riemannian metric since it is easy to compute and is widely used in applications. By using the Log-Euclidean distance in the loss function, we derive a shrinkage estimator in an analytic form and show that it is asymptotically optimal within a large class of estimators including the MLE, which is the sample Fr'echet mean of the data. We demonstrate the performance of the proposed shrinkage estimator via several simulated data experiments. Furthermore, we apply the shrinkage estimator to perform statistical inference in diffusion magnetic resonance imaging problems.

Full work available at URL: https://arxiv.org/abs/2007.02153

zbMATH Keywords

Stein's unbiased risk estimateFréchet meanTweedie's estimator

Mathematics Subject Classification ID

Statistics (62-XX)

Cites Work

• Estimation of the mean of a multivariate normal distribution
• A Limited Memory Algorithm for Bound Constrained Optimization
• Intrinsic statistics on Riemannian manifolds: Basic tools for geometric measurements
• Title not available (Why is that?)
• Title not available (Why is that?)
• Riemannian \(L^{p}\) center of mass: existence, uniqueness, and convexity
• Estimation with quadratic loss.
• Geometric Means in a Novel Vector Space Structure on Symmetric Positive‐Definite Matrices
• Using specially designed exponential families for density estimation
• Tweedie’s Formula and Selection Bias
• Title not available (Why is that?)
• Title not available (Why is that?)
• A Differential Geometric Approach to the Geometric Mean of Symmetric Positive-Definite Matrices
• Statistics on the manifold of multivariate normal distributions: theory and application to diffusion tensor MRI processing
• Shrinkage Estimators for Covariance Matrices
• Simultaneous Estimation of the Means of Independent Poisson Laws
• SURE Estimates for a Heteroscedastic Hierarchical Model
• Stein's Estimation Rule and Its Competitors--An Empirical Bayes Approach
• Data Mining and Knowledge Discovery Handbook
• Newton's method, zeroes of vector fields, and the Riemannian center of mass
• Optimal shrinkage of eigenvalues in the spiked covariance model
• Improving on inadmissible estimators in continuous exponential families with applications to simultaneous estimation of gamma scale parameters
• Regularizing flows for constrained matrix-valued images
• Non-Euclidean statistics for covariance matrices, with applications to diffusion tensor imaging
• Empirical null and false discovery rate inference for exponential families
• Minimizing a differentiable function over a differential manifold
• Posterior contraction rates of density derivative estimation
• Title not available (Why is that?)
• Weighted inductive means
• Estimation of noncentrality parameters
• Group Comparison of Eigenvalues and Eigenvectors of Diffusion Tensors
• Direct Density Derivative Estimation
• Statistics on the Stiefel manifold: theory and applications
• Harmonic Analysis on Symmetric Spaces—Higher Rank Spaces, Positive Definite Matrix Space and Generalizations
• Positive definite matrices: data representation and applications to computer vision
• Lognormal Distributions and Geometric Averages of Symmetric Positive Definite Matrices
• SURE estimates under dependence and heteroscedasticity
• An Empirical Bayes Method for Chi-Squared Data

Cited In (1)

• Diffusion means in geometric spaces

This page was built for publication: An Empirical Bayes Approach to Shrinkage Estimation on the Manifold of Symmetric Positive-Definite Matrices

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6153989)