Some density theorems in neural network with variable exponent
Mitsuo Izuki, Takahiro Noi, Yoshihiro Sawano, Hirokazu Tanaka
TL;DR
Several approximation theorems, originally formulated in the context of the standard Lp, are extended to the more general framework of variable exponent spaces, motivated by applications in neural networks.
Abstract
In this paper, we extend several approximation theorems, originally formulated in the context of the standard $L^p$ norm, to the more general framework of variable exponent spaces. Our study is motivated by applications in neural networks, where function approximation plays a crucial role. In addition to these generalizations, we provide alternative proofs for certain well-known results concerning the universal approximation property. In particular, we highlight spaces with variable exponents as illustrative examples, demonstrating the broader applicability of our approach.
