Core Concepts
G-Invariant and antisymmetric functions can be uniformly approximated by polynomials, with implications for deep learning.
Abstract
The content discusses the uniform Ck approximation of G-invariant and antisymmetric functions using polynomials. It explores the embedding dimensions, polynomial representations, and implications for deep learning. The article delves into the challenges of deep learning, the curse of dimensionality, and strategies to mitigate it. It also highlights the importance of symmetries in neural networks and their applications in various fields. The study focuses on approximations of G-invariant, symmetric, and antisymmetric functions due to their prevalence in science and technology.
Stats
G-invariant functions can be approximated by G-invariant polynomials.
Antisymmetric functions can be approximated as a sum of terms.
Embedding dimension is independent of regularity and accuracy.
Quotes
"For any subgroup G of the symmetric group Sn on n symbols, we present results for the uniform Ck approximation of G-invariant functions by G-invariant polynomials."
"We show that the embedding dimension required is independent of the regularity of the target function, the accuracy of the desired approximation, as well as k."