Geometry-aware principal component analysis for symmetric positive definite matrices

DOI10.1007/s10994-016-5605-5zbMath1459.62236OpenAlexW2554291319MaRDI QIDQ2398090

Inbal Horev, Masashi Sugiyama, Florian Yger

Publication date: 15 August 2017

Published in: Machine Learning (Search for Journal in Brave)

Full work available at URL: https://doi.org/10.1007/s10994-016-5605-5

zbMATH Keywords

Riemannian geometryGrassmann manifolddimensionality reductionPCAsymmetric positive definite (SPD) manifold

Mathematics Subject Classification ID

Factor analysis and principal components; correspondence analysis (62H25) Statistics on manifolds (62R30) Image analysis in multivariate analysis (62H35)

Related Items

Sion’s Minimax Theorem in Geodesic Metric Spaces and a Riemannian Extragradient Algorithm, Minimum cost‐compression risk in principal component analysis, Riemannian Hamiltonian Methods for Min-Max Optimization on Manifolds, Unsupervised manifold learning with polynomial mapping on symmetric positive definite matrices, Faster Riemannian Newton-type optimization by subsampling and cubic regularization, Goal-Oriented Optimal Approximations of Bayesian Linear Inverse Problems, cCorrGAN: conditional correlation GAN for learning empirical conditional distributions in the elliptope

Uses Software

• Manopt
• PRMLT

Cites Work

• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Log-determinant divergences revisited: alpha-beta and gamma log-det divergences
• The implicit function theorem. History, theory, and applications
• Non-Euclidean statistics for covariance matrices, with applications to diffusion tensor imaging
• Principal component analysis.
• Positive definite matrices
• A Riemannian framework for tensor computing
• Positive definite matrices and the S-divergence
• Manopt, a Matlab toolbox for optimization on manifolds
• Improved Estimation of Eigenvalues and Eigenvectors of Covariance Matrices Using Their Sample Estimates
• Computing the Fréchet Derivative of the Matrix Exponential, with an Application to Condition Number Estimation
• The Geometry of Algorithms with Orthogonality Constraints
• From Manifold to Manifold: Geometry-Aware Dimensionality Reduction for SPD Matrices
• The Riemannian Mean of Positive Matrices
• Stochastic Gradient Descent on Riemannian Manifolds
• Geometric Means in a Novel Vector Space Structure on Symmetric Positive‐Definite Matrices
• Asymptotic Theory for Principal Component Analysis