A note on Cybenko's Universal Approximation Theorem

2025/08/26 by Wang, Kun
#Classical Analysis and ODEs (math.CA) #FOS: Computer and information sciences #FOS: Mathematics #Neural and Evolutionary Computing (cs.NE)

paper · doi:10.48550/arxiv.2508.18893

Abstract

In this short note, we point out a mistake in G.Cybenko's proof of his version of the universal approximation theorem which has been widely cited. This mistake might not be easily fixable along the idea of his proof and it also leads to an interesting question in measure theory.

Citations

• Multilayer feedforward networks with a nonpolynomial activation function can approximate any function

Related

• Theoreme d'approximation cylindrique dans le cas des anneaux analytiques
• Corrections to "Universal Densities Exist for Every Finite Reference Measure"
• A note on Kronecker's approximation theorem
• Approximations to the Kruskal-Katona theorem
• On approximation schemes and compactness