Think globally, fit locally under the manifold setup: asymptotic analysis of locally linear embedding

DOI10.1214/17-AOS1676zbMATH Open1405.62058arXiv1703.04058OpenAlexW2595579230WikidataQ125835170 ScholiaQ125835170MaRDI QIDQ1990602FDOQ1990602

H.-T. Wu, Nan Wu

Publication date: 25 October 2018

Published in: The Annals of Statistics (Search for Journal in Brave)

Abstract: Since its introduction in 2000, the locally linear embedding (LLE) has been widely applied in data science. We provide an asymptotical analysis of the LLE under the manifold setup. We show that for the general manifold, asymptotically we may not obtain the Laplace-Beltrami operator, and the result may depend on the non-uniform sampling, unless a correct regularization is chosen. We also derive the corresponding kernel function, which indicates that the LLE is not a Markov process. A comparison with the other commonly applied nonlinear algorithms, particularly the diffusion map, is provided, and its relationship with the locally linear regression is also discussed.

Full work available at URL: https://arxiv.org/abs/1703.04058

Recommendations

• 10.1162/153244304322972667
• The LLE and a linear mapping
• scientific article; zbMATH DE number 2013268
• Application of local linear embedding to nonlinear exploratory latent structure analysis
• Dependence of locally linear embedding on the regularization parameter

zbMATH Keywords

dimension reductionmeasurement errordiffusion mapslocally linear embeddinglocally linear regression

Mathematics Subject Classification ID

Nonparametric estimation (62G05) Linear regression; mixed models (62J05) Learning and adaptive systems in artificial intelligence (68T05) Estimation in multivariate analysis (62H12) Partial differential equations on manifolds; differential operators (58J99)

Cites Work

• Title not available (Why is that?)
• Diffusion maps
• Title not available (Why is that?)
• Consistency of spectral clustering
• Title not available (Why is that?)
• Title not available (Why is that?)
• Embedding Riemannian manifolds by their heat kernel
• Laplacian Eigenmaps for Dimensionality Reduction and Data Representation
• Chernoff's theorem and discrete time approximations of Brownian motion on manifolds
• Finite propagation speed, kernel estimates for functions of the Laplace operator, and the geometry of complete Riemannian manifolds
• From graph to manifold Laplacian: the convergence rate
• Vector diffusion maps and the connection Laplacian
• Empirical graph Laplacian approximation of Laplace–Beltrami operators: Large sample results
• Principal Manifolds and Nonlinear Dimensionality Reduction via Tangent Space Alignment
• Hessian eigenmaps: Locally linear embedding techniques for high-dimensional data
• Learning Theory
• Optimal shrinkage of eigenvalues in the spiked covariance model
• Spectral geometry: direct and inverse problems. With an appendix by G. Besson
• Graph connection Laplacian methods can be made robust to noise
• On information plus noise kernel random matrices
• Learning Theory
• The strong uniform consistency of nearest neighbor density estimates
• Consistency properties of nearest neighbor density function estimators
• Spectral convergence of the connection Laplacian from random samples
• A variational approach to the consistency of spectral clustering
• Two-dimensional tomography from noisy projections taken at unknown random directions
• Local Linear Regression on Manifolds and Its Geometric Interpretation

Cited In (10)

• Eigen-convergence of Gaussian kernelized graph Laplacian by manifold heat interpolation
• Data-driven efficient solvers for Langevin dynamics on manifold in high dimensions
• Time-series forecasting using manifold learning, radial basis function interpolation, and geometric harmonics
• Rates of the strong uniform consistency with rates for conditional \(U\)-statistics estimators with general kernels on manifolds
• Diffusion maps for embedded manifolds with boundary with applications to PDEs
• Learning low-dimensional nonlinear structures from high-dimensional noisy data: an integral operator approach
• Connecting dots: from local covariance to empirical intrinsic geometry and locally linear embedding
• Embeddings of Riemannian manifolds with finite eigenvector fields of connection Laplacian
• Spectral convergence of graph Laplacian and heat kernel reconstruction in \(L^\infty\) from random samples
• Rates of the strong uniform consistency for the kernel-type regression function estimators with general kernels on manifolds

Uses Software

• t-SNE

This page was built for publication: Think globally, fit locally under the manifold setup: asymptotic analysis of locally linear embedding

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1990602)