Positivity sets of hinge functions

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Schicho, Josef, Tewari, Ayush Kumar, Warren, Audie
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912279541317632
author Schicho, Josef
Tewari, Ayush Kumar
Warren, Audie
author_facet Schicho, Josef
Tewari, Ayush Kumar
Warren, Audie
contents In this paper we investigate which subsets of the real plane are realisable as the set of points on which a one-layer ReLU neural network takes a positive value. In the case of cones we give a full characterisation of such sets. Furthermore, we give a necessary condition for any subset of $\mathbb R^d$. We give various examples of such one-layer neural networks.
format Preprint
id arxiv_https___arxiv_org_abs_2503_13512
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Positivity sets of hinge functions
Schicho, Josef
Tewari, Ayush Kumar
Warren, Audie
Machine Learning
Discrete Mathematics
Symbolic Computation
Combinatorics
Functional Analysis
In this paper we investigate which subsets of the real plane are realisable as the set of points on which a one-layer ReLU neural network takes a positive value. In the case of cones we give a full characterisation of such sets. Furthermore, we give a necessary condition for any subset of $\mathbb R^d$. We give various examples of such one-layer neural networks.
title Positivity sets of hinge functions
topic Machine Learning
Discrete Mathematics
Symbolic Computation
Combinatorics
Functional Analysis
url https://arxiv.org/abs/2503.13512