Cargando…

Optimizing the Simplicial-Map Neural Network Architecture

Simplicial-map neural networks are a recent neural network architecture induced by simplicial maps defined between simplicial complexes. It has been proved that simplicial-map neural networks are universal approximators and that they can be refined to be robust to adversarial attacks. In this paper,...

Descripción completa

Detalles Bibliográficos
Autores principales: Paluzo-Hidalgo, Eduardo, Gonzalez-Diaz, Rocio, Gutiérrez-Naranjo, Miguel A., Heras, Jónathan
Formato: Online Artículo Texto
Lenguaje:English
Publicado: MDPI 2021
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8466576/
https://www.ncbi.nlm.nih.gov/pubmed/34564099
http://dx.doi.org/10.3390/jimaging7090173
_version_ 1784573176581193728
author Paluzo-Hidalgo, Eduardo
Gonzalez-Diaz, Rocio
Gutiérrez-Naranjo, Miguel A.
Heras, Jónathan
author_facet Paluzo-Hidalgo, Eduardo
Gonzalez-Diaz, Rocio
Gutiérrez-Naranjo, Miguel A.
Heras, Jónathan
author_sort Paluzo-Hidalgo, Eduardo
collection PubMed
description Simplicial-map neural networks are a recent neural network architecture induced by simplicial maps defined between simplicial complexes. It has been proved that simplicial-map neural networks are universal approximators and that they can be refined to be robust to adversarial attacks. In this paper, the refinement toward robustness is optimized by reducing the number of simplices (i.e., nodes) needed. We have shown experimentally that such a refined neural network is equivalent to the original network as a classification tool but requires much less storage.
format Online
Article
Text
id pubmed-8466576
institution National Center for Biotechnology Information
language English
publishDate 2021
publisher MDPI
record_format MEDLINE/PubMed
spelling pubmed-84665762021-10-28 Optimizing the Simplicial-Map Neural Network Architecture Paluzo-Hidalgo, Eduardo Gonzalez-Diaz, Rocio Gutiérrez-Naranjo, Miguel A. Heras, Jónathan J Imaging Article Simplicial-map neural networks are a recent neural network architecture induced by simplicial maps defined between simplicial complexes. It has been proved that simplicial-map neural networks are universal approximators and that they can be refined to be robust to adversarial attacks. In this paper, the refinement toward robustness is optimized by reducing the number of simplices (i.e., nodes) needed. We have shown experimentally that such a refined neural network is equivalent to the original network as a classification tool but requires much less storage. MDPI 2021-09-01 /pmc/articles/PMC8466576/ /pubmed/34564099 http://dx.doi.org/10.3390/jimaging7090173 Text en © 2021 by the authors. https://creativecommons.org/licenses/by/4.0/Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
spellingShingle Article
Paluzo-Hidalgo, Eduardo
Gonzalez-Diaz, Rocio
Gutiérrez-Naranjo, Miguel A.
Heras, Jónathan
Optimizing the Simplicial-Map Neural Network Architecture
title Optimizing the Simplicial-Map Neural Network Architecture
title_full Optimizing the Simplicial-Map Neural Network Architecture
title_fullStr Optimizing the Simplicial-Map Neural Network Architecture
title_full_unstemmed Optimizing the Simplicial-Map Neural Network Architecture
title_short Optimizing the Simplicial-Map Neural Network Architecture
title_sort optimizing the simplicial-map neural network architecture
topic Article
url https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8466576/
https://www.ncbi.nlm.nih.gov/pubmed/34564099
http://dx.doi.org/10.3390/jimaging7090173
work_keys_str_mv AT paluzohidalgoeduardo optimizingthesimplicialmapneuralnetworkarchitecture
AT gonzalezdiazrocio optimizingthesimplicialmapneuralnetworkarchitecture
AT gutierreznaranjomiguela optimizingthesimplicialmapneuralnetworkarchitecture
AT herasjonathan optimizingthesimplicialmapneuralnetworkarchitecture