Robust geodesic regression

DOI10.48550/ARXIV.2007.04518zbMATH Open1490.62514arXiv2007.04518OpenAlexW4205826957MaRDI QIDQ122252FDOQ122252

Authors: Ha-Young Shin, Hee-Seok Oh, Ha-Young Shin, Hee-Seok Oh 

Publication date: 9 July 2020

Published in: International Journal of Computer Vision (Search for Journal in Brave)

Abstract: This paper studies robust regression for data on Riemannian manifolds. Geodesic regression is the generalization of linear regression to a setting with a manifold-valued dependent variable and one or more real-valued independent variables. The existing work on geodesic regression uses the sum-of-squared errors to find the solution, but as in the classical Euclidean case, the least-squares method is highly sensitive to outliers. In this paper, we use M-type estimators, including the $L_1$, Huber and Tukey biweight estimators, to perform robust geodesic regression, and describe how to calculate the tuning parameters for the latter two. We also show that, on compact symmetric spaces, all M-type estimators are maximum likelihood estimators, and argue for the overall superiority of the $L_1$ estimator over the $L_2$ and Huber estimators on high-dimensional manifolds and over the Tukey biweight estimator on compact high-dimensional manifolds. Results from numerical examples, including analysis of real neuroimaging data, demonstrate the promising empirical properties of the proposed approach.

Full work available at URL: https://arxiv.org/abs/2007.04518

Recommendations

• Geodesic regression and the theory of least squares on Riemannian manifolds
• Robust nonparametric regression on Riemannian manifolds
• Probabilistic geodesic models for regression and dimensionality reduction on Riemannian manifolds
• Geodesic regression and its application to shape analysis
• Regression models on Riemannian symmetric spaces

zbMATH Keywords

robust statisticsgeodesic regressionM-type estimatorsmanifold statisticsRiemannian manifolds

Mathematics Subject Classification ID

General nonlinear regression (62J02) Statistics on manifolds (62R30) Methods of local Riemannian geometry (53B21)

Cites Work

• Regression models on Riemannian symmetric spaces
• The double bubble problem in spherical and hyperbolic space
• Geodesic regression and the theory of least squares on Riemannian manifolds
• Intrinsic polynomials for regression on Riemannian manifolds
• A novel dynamic system in the space of SPD matrices with applications to appearance tracking
• Title not available (Why is that?)
• Completely monotone functions and the Wallis ratio
• Statistics on manifolds
• Nonparametric Regression between General Riemannian Manifolds

Cited In (10)

• Regression models on Riemannian symmetric spaces
• Regression of fluctuating system properties: baryonic tully-Fisher scaling in disk galaxies
• Probabilistic geodesic models for regression and dimensionality reduction on Riemannian manifolds
• Robust nonparametric regression on Riemannian manifolds
• Geodesic regression on spheres from a numerical optimization viewpoint
• Geodesic regression and its application to shape analysis
• Title not available (Why is that?)
• Interacting Geometric Priors For Robust Multimodel Fitting
• Geodesic least squares regression on the Gaussian manifold with an application in astrophysics
• GeodRegr

This page was built for publication: Robust geodesic regression

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q122252)