Geometry of Polynomial Neural Networks

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Kubjas, Kaie, Li, Jiayi, Wiesmann, Maximilian
Format: Preprint
Veröffentlicht: 2024
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866909413678252032
author Kubjas, Kaie
Li, Jiayi
Wiesmann, Maximilian
author_facet Kubjas, Kaie
Li, Jiayi
Wiesmann, Maximilian
contents We study the expressivity and learning process for polynomial neural networks (PNNs) with monomial activation functions. The weights of the network parametrize the neuromanifold. In this paper, we study certain neuromanifolds using tools from algebraic geometry: we give explicit descriptions as semialgebraic sets and characterize their Zariski closures, called neurovarieties. We study their dimension and associate an algebraic degree, the learning degree, to the neurovariety. The dimension serves as a geometric measure for the expressivity of the network, the learning degree is a measure for the complexity of training the network and provides upper bounds on the number of learnable functions. These theoretical results are accompanied with experiments.
format Preprint
id arxiv_https___arxiv_org_abs_2402_00949
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Geometry of Polynomial Neural Networks
Kubjas, Kaie
Li, Jiayi
Wiesmann, Maximilian
Algebraic Geometry
Machine Learning
68T07, 14P10, 14N07, 14M12
We study the expressivity and learning process for polynomial neural networks (PNNs) with monomial activation functions. The weights of the network parametrize the neuromanifold. In this paper, we study certain neuromanifolds using tools from algebraic geometry: we give explicit descriptions as semialgebraic sets and characterize their Zariski closures, called neurovarieties. We study their dimension and associate an algebraic degree, the learning degree, to the neurovariety. The dimension serves as a geometric measure for the expressivity of the network, the learning degree is a measure for the complexity of training the network and provides upper bounds on the number of learnable functions. These theoretical results are accompanied with experiments.
title Geometry of Polynomial Neural Networks
topic Algebraic Geometry
Machine Learning
68T07, 14P10, 14N07, 14M12
url https://arxiv.org/abs/2402.00949