Optimization over geodesics for exact principal geodesic analysis

DOI10.1007/S10444-013-9308-1zbMATH Open1300.65045arXiv1008.1902OpenAlexW2099327136MaRDI QIDQ457684FDOQ457684

Mads Nielsen, Stefan Sommer, François Lauze

Publication date: 29 September 2014

Published in: Advances in Computational Mathematics (Search for Journal in Brave)

Abstract: In fields ranging from computer vision to signal processing and statistics, increasing computational power allows a move from classical linear models to models that incorporate non-linear phenomena. This shift has created interest in computational aspects of differential geometry, and solving optimization problems that incorporate non-linear geometry constitutes an important computational task. In this paper, we develop methods for numerically solving optimization problems over spaces of geodesics using numerical integration of Jacobi fields and second order derivatives of geodesic families. As an important application of this optimization strategy, we compute exact Principal Geodesic Analysis (PGA), a non-linear version of the PCA dimensionality reduction procedure. By applying the exact PGA algorithm to synthetic data, we exemplify the differences between the linearized and exact algorithms caused by the non-linear geometry. In addition, we use the numerically integrated Jacobi fields to determine sectional curvatures and provide upper bounds for injectivity radii.

Full work available at URL: https://arxiv.org/abs/1008.1902

zbMATH Keywords

dimensionality reductionnumerical examplesalgorithmmanifold statisticsgeometric optimizationdifferential geometryRiemannian metricsprincipal geodesic analysis

Mathematics Subject Classification ID

Factor analysis and principal components; correspondence analysis (62H25) Numerical optimization and variational techniques (65K10) Existence theories for optimal control problems involving ordinary differential equations (49J15) Differential topology (57R99)

Cites Work

• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• MANPAK: A set of algorithms for computations on implicitly defined manifolds
• Shape Manifolds, Procrustean Metrics, and Complex Projective Spaces
• Intrinsic statistics on Riemannian manifolds: Basic tools for geometric measurements
• The Differentiation of Pseudo-Inverses and Nonlinear Least Squares Problems Whose Variables Separate
• Geodesic active contours
• Riemannian center of mass and mollifier smoothing
• A Riemannian framework for tensor computing
• Riemannian geometry for the statistical analysis of diffusion tensor data
• The Gradient Projection Method Along Geodesics
• Symplectic methods for the approximation of the exponential map and the Newton iteration on Riemannian submanifolds
• Extensions and Applications of the Householder Algorithm for Solving Linear Least Squares Problems
• On the derivative of the generalized inverse of a matrix
• Horizontal Dimensionality Reduction and Iterated Frame Bundle Development
• Curvature computations for \(n\)-manifolds in \(\mathbb{R}^{n+m}\) and solution to an open problem proposed by R. Goldman
• On a computational method for the second fundamental tensor and its application to bifurcation problems
• Globally convergent optimization algorithms on Riemannian manifolds: Uniform framework for unconstrained and constrained optimization

Cited In (13)

• On the robust PCA and Weiszfeld's algorithm
• Principal component geodesics for planar shape spaces
• 3D human pose tracking priors using geodesic mixture models
• Barycentric subspace analysis on manifolds
• Long-time principal geodesic analysis in director-based dynamics of hybrid mechanical systems
• Recent advances in directional statistics
• Localizing differentially evolving covariance structures via scan statistics
• Nonlinear model reduction on metric spaces. Application to one-dimensional conservative PDEs in Wasserstein spaces
• Canonical Correlation Analysis on SPD(n) Manifolds
• Point-Process Principal Components Analysis via Geometric Optimization
• Principal symmetric space analysis
• Sparse Wasserstein barycenters and application to reduced order modeling
• Advances in Geometric Statistics for Manifold Dimension Reduction

Uses Software

• MANPAK

This page was built for publication: Optimization over geodesics for exact principal geodesic analysis

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q457684)