Fully hyperbolic convolutional neural networks

DOI10.1007/S40687-022-00343-1MaRDI QIDQ2079530zbMATH OpenOpenAlexWikidataFDO

Authors Keegan Lensink, Bas Peters, Eldad Haber

Publication date 30 September 2022

Published in Research in the Mathematical Sciences (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1905.10484

zbMATH Keywords

waveletsneural networkshyperbolic PDEauto-encoders

Mathematics Subject Classification ID

Computer science (68-XX) Information and communication theory, circuits (94-XX)

Recommendations

• Convolutional Neural Networks in Phase Space and Inverse Problems
• Theory of deep convolutional neural networks. II: Spherical analysis
• Homogeneous alternating projection neural networks
• Multivariate hyperbolic tangent neural network approximation
• Convolution hierarchical deep-learning neural networks (C-HiDeNN): finite elements, isogeometric analysis, tensor decomposition, and beyond
• Equivalence of approximation by convolutional neural networks and fully-connected networks
• Exploring hypersurfaces with offset-like convolutions

Cites work

• scientific article; zbMATH DE number 236854 (Why is no real title available?)
• A Crank-Nicolson collocation spectral method for the two-dimensional telegraph equations
• Deep learning
• Deep neural networks motivated by partial differential equations
• Group invariant scattering
• Learning deep architectures for AI
• Orthonormal bases of compactly supported wavelets

Cited in

(3)

• Estimating a Potential Without the Agony of the Partition Function
• scientific article; zbMATH DE number 67881 (Why is no real title available?)
• Hyper-flexible convolutional neural networks based on generalized Lehmer and power means

Describes a project that uses

Uses Software

• StyleGAN
• PacGAN
• BigGAN
• VEEGAN
• ANODE
• AlexNet
• GitHub
• U-Net
• i-RevNet
• ImageNet
• PyTorch
• MNIST

This page was built for publication: Fully hyperbolic convolutional neural networks

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2079530)