Do ideas have shape? Idea registration as the continuous limit of artificial neural networks

DOI10.1016/J.PHYSD.2022.133592OpenAlexW4309879846MaRDI QIDQ2111734FDOQ2111734

Publication date: 17 January 2023

Published in: Physica D (Search for Journal in Brave)

Full work available at URL: https://arxiv.org/abs/2008.03920

Recommendations

The emergence of a concept in shallow neural networks
Abstract neural automata: variability of structure, thought and Riemannian volume
Cognition: Differential-geometrical view on neural networks
Continuous functions and neural network semantics
Feed-forward neural networks: a geometrical perspective
On fixed points of continuous mappings associated with construction of artificial neural networks
On neural network architecture based on concept lattices
Neural ideals and stimulus space visualization
Characterization of the variation spaces corresponding to shallow neural networks
A formalisation of the neural assembly concept. I: Constraints on neural assembly size

regularization image registration deep Gaussian processes residual neural network Hamiltonian and Lagrangian analysis infinite depth limit

Mathematics Subject Classification ID

Numerical analysis (65-XX) Computer science (68-XX)

Cites Work

Computing large deformation metric mappings via geodesic flows of diffeomorphisms
Title not available (Why is that?)
Universal approximation bounds for superpositions of a sigmoidal function
Computable Elastic Distances Between Shapes
Bayesian Inference for Non-Stationary Spatial Covariance Structure via Spatial Deformations
Shapes and diffeomorphisms
Kernels for vector-valued functions: a review
Multigrid with Rough Coefficients and Multiresolution Operator Decomposition from Hierarchical Information Games
Support Vector Machines
Operator-Adapted Wavelets, Fast Solvers, and Numerical Homogenization
Probabilistic numerics and uncertainty in computations
Compression, inversion, and approximate PCA of dense kernel matrices at near-linear computational complexity
Title not available (Why is that?)
Discrete mechanics and variational integrators
Title not available (Why is that?)
Geometric Numerical Integration
Metric-based upscaling
Bayesian Numerical Homogenization
Approximation of the effective conductivity of ergodic media by periodization
Title not available (Why is that?)
On completeness of groups of diffeomorphisms
Geometric numerical integration illustrated by the Störmer–Verlet method
Variational problems on flows of diffeomorphisms for image matching
Title not available (Why is that?)
The Euler-Poincaré theory of metamorphosis
On the \(\Gamma\)-convergence of discrete dynamics and variational integrators
Landmark matching via large deformation diffeomorphisms
Local error estimates for radial basis function interpolation of scattered data
Metamorphoses through Lie group action
The momentum map representation of images
Geodesic shooting for computational anatomy
Computational anatomy: an emerging discipline
Qualitative Robustness in Bayesian Inference
Diffeomorphic 3D image registration via geodesic shooting using an efficient adjoint calculation
The regularizing properties of the composite trapezoidal method for weakly singular Volterra integral equations of the first kind
Strong solutions for differential equations in abstract spaces
Title not available (Why is that?)
Sectional curvature in terms of the cometric, with applications to the Riemannian manifolds of landmarks
Operator-valued kernels for learning from functional response data
Kernel Mode Decomposition and the Programming of Kernels
Existence of solutions for ordinary differential equations in Banach spaces
A concise introduction to geometric numerical integration
Brittleness of Bayesian inference under finite information in a continuous world
On the brittleness of Bayesian inference
Brittleness of Bayesian inference and new Selberg formulas
Physics-informed neural networks: a deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations
A mean-field optimal control formulation of deep learning
Bayesian Probabilistic Numerical Methods
Metamorphoses of functional shapes in Sobolev spaces
Stable architectures for deep neural networks
Shapes and diffeomorphisms
A proposal on machine learning via dynamical systems
Title not available (Why is that?)
A mean field view of the landscape of two-layer neural networks
PCANet: A Simple Deep Learning Baseline for Image Classification?
Geodesic Image Regression with a Sparse Parameterization of Diffeomorphisms
Kernel flows: from learning kernels from data into the abyss
Deep regularization and direct training of the inner layers of neural networks with kernel flows
Learning dynamical systems from data: a simple cross-validation perspective. I: Parametric kernel flows
Statistical Numerical Approximation
Learning ``best kernels from data in Gaussian process regression. With application to aerodynamics
Uncertainty quantification of the 4th kind; optimal posterior accuracy-uncertainty tradeoff with the minimum enclosing ball
Solving and learning nonlinear PDEs with Gaussian processes
The Random Feature Model for Input-Output Maps between Banach Spaces
Consistency of empirical Bayes and kernel flow for hierarchical parameter estimation
Sparse Cholesky Factorization by Kullback--Leibler Minimization
Hyperparameter estimation in Bayesian MAP estimation: parameterizations and consistency
Computational graph completion
A representer theorem for deep kernel learning

Cited In (6)

Operator learning using random features: a tool for scientific computing
Gaussian process hydrodynamics
Learning dynamical systems from data: a simple cross-validation perspective. IV: Case with partial observations
Kernel methods are competitive for operator learning
Error analysis of kernel/GP methods for nonlinear and parametric PDEs
Operator learning with Gaussian processes

Uses Software

LDDMM

This page was built for publication: Do ideas have shape? Idea registration as the continuous limit of artificial neural networks

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2111734)