Deep nonparametric estimation of intrinsic data structures by chart autoencoders: generalization error and robustness

DOI10.1016/j.acha.2023.101602arXiv2303.09863OpenAlexW4387580637MaRDI QIDQ6144890

Unnamed Author, Alex Havrilla, Rongjie Lai, Wenjing Liao

Publication date: 30 January 2024

Published in: Applied and Computational Harmonic Analysis (Search for Journal in Brave)

Full work available at URL: https://arxiv.org/abs/2303.09863

zbMATH Keywords

dimension reductiongeneralization errordeep learning theorymanifold modelchart autoencoder

Mathematics Subject Classification ID

Statistics (62-XX) Computer science (68-XX)

Cites Work

• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Multi-scale geometric methods for data sets. II: Geometric multi-resolution analysis
• Manifold estimation and singular deconvolution under Hausdorff loss
• Auto-association by multilayer perceptrons and singular value decomposition
• Multi-fidelity Gaussian process regression for prediction of random fields
• Multilayer feedforward networks are universal approximators
• Nonasymptotic rates for manifold, tangent space and curvature estimation
• Provable approximation properties for deep neural networks
• A distribution-free theory of nonparametric regression
• Weak convergence and empirical processes. With applications to statistics
• Error bounds for deep ReLU networks using the Kolmogorov-Arnold superposition theorem
• Adaptive two-layer ReLU neural network. I: Best least-squares approximation
• Error bounds for approximations with deep ReLU networks
• The reach, metric distortion, geodesic convexity and the variation of tangent spaces
• Finding the homology of submanifolds with high confidence from random samples
• Side effects of learning from low-dimensional data embedded in a Euclidean space
• Accurate Uncertainty Quantification Using Inaccurate Computational Models
• Curvature Measures
• Learning Deep Architectures for AI
• Geometric diffusions as a tool for harmonic analysis and structure definition of data: Diffusion maps
• Laplacian Eigenmaps for Dimensionality Reduction and Data Representation
• Adaptive Geometric Multiscale Approximations for Intrinsically Low-dimensional Data
• Minimax Manifold Estimation
• Manifold reconstruction using tangential Delaunay complexes
• Testing the manifold hypothesis
• A deep network construction that adapts to intrinsic dimensionality beyond the domain

This page was built for publication: Deep nonparametric estimation of intrinsic data structures by chart autoencoders: generalization error and robustness