An unfeasibility view of neural network learning

DOI10.1016/J.JCO.2022.101710MaRDI QIDQ2685066zbMATH OpenOpenAlexWikidataFDO

Authors Joos Heintz, Enrique Carlos Segura, Hvara Ocar, Andrés Rojas Paredes, Luis Miguel Pardo

Publication date 17 February 2023

Published in Journal of Complexity (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/2201.00945

zbMATH Keywords

activation functioncomplexity lower boundcontinuously differentiable functionepicyclemachine learningmultilayer neural network

Mathematics Subject Classification ID

Learning and adaptive systems in artificial intelligence (68T05)

Recommendations

• scientific article; zbMATH DE number 774006
• On the infeasibility of training neural networks with small mean-squared error
• scientific article; zbMATH DE number 1784876
• The computational intractability of training sigmoidal neural networks
• The learning problem of multi-layer neural networks
• Unlearning in the paramagnetic phase of neural network models
• Learning in the feed-forward random neural network: A critical review
• scientific article; zbMATH DE number 774004
• Neural Network Learning as an Inverse Problem

Cites work

• Complexity of cylindrical decompositions of sub-Pfaffian
• scientific article; zbMATH DE number 6000883 (Why is no real title available?)
• Multilayer feedforward networks are universal approximators
• Neural Network Learning
• On the computational complexity and geometry of the first-order theory of the reals. I: Introduction. Preliminaries. The geometry of semi-algebraic sets. The decision problem for the existential theory of the reals
• Quiz games as a model for information hiding
• Sur la complexité du principe de Tarski-Seidenberg

Cited in

(11)

• Learning what we don't care about: anti-training with sacrificial functions
• scientific article; zbMATH DE number 1304387 (Why is no real title available?)
• Learning the unlearnable
• Distribution-specific hardness of learning neural networks
• Monotone learning with rectified wire networks
• Neural networks using a logistics sigmoid function: linear classifier bounds and global nonattainability
• Can We Teach Functions to an Artificial Intelligence by Just Showing It Enough “Ground Truth”?
• Solution Existence Theory for Artificial Neural Networks
• In memoriam. Succinct obituary in memoriam of Joos Heintz
• Dedicated to the memory of Joos Heintz
• Erzeugungsgrad, VC-dimension and neural networks with rational activation function

This page was built for publication: An unfeasibility view of neural network learning

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2685066)