Deep Networks are Reproducing Kernel Chains

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Heeringa, Tjeerd Jan, Spek, Len, Brune, Christoph
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866929663517917184
author Heeringa, Tjeerd Jan
Spek, Len
Brune, Christoph
author_facet Heeringa, Tjeerd Jan
Spek, Len
Brune, Christoph
contents Identifying an appropriate function space for deep neural networks remains a key open question. While shallow neural networks are naturally associated with Reproducing Kernel Banach Spaces (RKBS), deep networks present unique challenges. In this work, we extend RKBS to chain RKBS (cRKBS), a new framework that composes kernels rather than functions, preserving the desirable properties of RKBS. We prove that any deep neural network function is a neural cRKBS function, and conversely, any neural cRKBS function defined on a finite dataset corresponds to a deep neural network. This approach provides a sparse solution to the empirical risk minimization problem, requiring no more than $N$ neurons per layer, where $N$ is the number of data points.
format Preprint
id arxiv_https___arxiv_org_abs_2501_03697
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Deep Networks are Reproducing Kernel Chains
Heeringa, Tjeerd Jan
Spek, Len
Brune, Christoph
Machine Learning
Functional Analysis
46E15 (Primary) 46B10, 68T07, 26B40 (Secondary)
I.2.6; G.1.6
Identifying an appropriate function space for deep neural networks remains a key open question. While shallow neural networks are naturally associated with Reproducing Kernel Banach Spaces (RKBS), deep networks present unique challenges. In this work, we extend RKBS to chain RKBS (cRKBS), a new framework that composes kernels rather than functions, preserving the desirable properties of RKBS. We prove that any deep neural network function is a neural cRKBS function, and conversely, any neural cRKBS function defined on a finite dataset corresponds to a deep neural network. This approach provides a sparse solution to the empirical risk minimization problem, requiring no more than $N$ neurons per layer, where $N$ is the number of data points.
title Deep Networks are Reproducing Kernel Chains
topic Machine Learning
Functional Analysis
46E15 (Primary) 46B10, 68T07, 26B40 (Secondary)
I.2.6; G.1.6
url https://arxiv.org/abs/2501.03697