On PDE characterization of smooth hierarchical functions computed by neural networks

DOI10.1162/NECO_A_01441MaRDI QIDQ5034481zbMATH OpenOpenAlexWikidataFDO

Authors Khashayar Filom, Roozbeh Farhoodi, Konrad P. Körding

Publication date 18 February 2022

Published in Neural Computation (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/2005.08859

Mathematics Subject Classification ID

Computer science (68-XX)

Recommendations

• Feedforward neural networks and compositional functions with applications to dynamical systems
• On functions computed on trees
• Approximation rates for neural networks with encodable weights in smoothness spaces
• Neural networks for optimal approximation of continuous functions in \(\mathbb R^d\)
• Provable approximation properties for deep neural networks

Cites work

• A universal differential equation
• Approximate complexity and functional representation
• Approximation by superpositions of a sigmoidal function
• Characterization of Classes of Functions
• From Hilbert's 13th problem to the theory of neural networks: constructive aspects of Kolmogorov's superposition theorem
• scientific article; zbMATH DE number 3147701 (Why is no real title available?)
• scientific article; zbMATH DE number 3758614 (Why is no real title available?)
• scientific article; zbMATH DE number 3551408 (Why is no real title available?)
• scientific article; zbMATH DE number 1949151 (Why is no real title available?)
• scientific article; zbMATH DE number 1749058 (Why is no real title available?)
• scientific article; zbMATH DE number 1765103 (Why is no real title available?)
• scientific article; zbMATH DE number 3234662 (Why is no real title available?)
• scientific article; zbMATH DE number 3246460 (Why is no real title available?)
• scientific article; zbMATH DE number 3259469 (Why is no real title available?)
• scientific article; zbMATH DE number 3284515 (Why is no real title available?)
• scientific article; zbMATH DE number 3300386 (Why is no real title available?)
• scientific article; zbMATH DE number 3314813 (Why is no real title available?)
• scientific article; zbMATH DE number 3086990 (Why is no real title available?)
• scientific article; zbMATH DE number 3096281 (Why is no real title available?)
• Multilayer feedforward networks are universal approximators
• On functions computed on trees
• On Hilbert's thirteenth problem and related questions
• On the Structure of Continuous Functions of Several Variables
• Spurious valleys in one-hidden-layer neural network optimization landscapes
• The solutions of a smooth PDE can be dense in C[I]
• Theoretical Insights Into the Optimization Landscape of Over-Parameterized Shallow Neural Networks
• Topological properties of the set of functions generated by neural networks of fixed size
• Why does deep and cheap learning work so well?

Cited in

(5)

• scientific article; zbMATH DE number 1164037 (Why is no real title available?)
• Feedforward neural networks and compositional functions with applications to dynamical systems
• On functions computed on trees
• PDE constraints on smooth hierarchical functions computed by neural networks
• Hyper-flexible convolutional neural networks based on generalized Lehmer and power means

This page was built for publication: On PDE characterization of smooth hierarchical functions computed by neural networks

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5034481)