Measure theoretic results for approximation by neural networks with limited weights

DOI10.1080/01630563.2016.1254654MaRDI QIDQ6432064arXivFDO

Authors Vugar E. Ismailov, Ekrem Savaş

Publication date 4 April 2023

Mathematics Subject Classification ID

Spaces of measures (46E27) Multidimensional problems (41A63) Neural networks for/in biological studies, artificial life and related topics (92B20) Spaces of measures, convergence of measures (28A33) Approximation by other special function classes (41A30)

Abstract: In this paper, we study approximation properties of single hidden layer neural networks with weights varying on finitely many directions and thresholds from an open interval. We obtain a necessary and at the same time sufficient measure theoretic condition for density of such networks in the space of continuous functions. Further, we prove a density result for neural networks with a specifically constructed activation function and a fixed number of neurons.

This page was built for publication: Measure theoretic results for approximation by neural networks with limited weights

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6432064)