VC dimensions of group convolutional neural networks

Autor:	Petersen, Philipp Christian, Sepliarskaia, Anna
Rok vydání:	2022
Předmět:	Computer Science - Machine Learning Mathematics - Functional Analysis Statistics - Machine Learning 68T07 68Q32 68T05
Druh dokumentu:	Working Paper
Popis:	We study the generalization capacity of group convolutional neural networks. We identify precise estimates for the VC dimensions of simple sets of group convolutional neural networks. In particular, we find that for infinite groups and appropriately chosen convolutional kernels, already two-parameter families of convolutional neural networks have an infinite VC dimension, despite being invariant to the action of an infinite group.
Databáze:	arXiv
Externí odkaz:	http://arxiv.org/abs/2212.09507 Zobrazit plný text záznamu View this record from Arxiv