On the Sample Complexity of Two-Layer Networks: Lipschitz vs. Element-Wise Lipschitz Activation

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Daniely, Amit, Granot, Elad
Format: Preprint
Veröffentlicht: 2022
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866917570874966016
author Daniely, Amit
Granot, Elad
author_facet Daniely, Amit
Granot, Elad
contents We investigate the sample complexity of bounded two-layer neural networks using different activation functions. In particular, we consider the class $$ \mathcal{H} = \left\{\textbf{x}\mapsto \langle \textbf{v}, σ\circ W\textbf{b} + \textbf{b} \rangle : \textbf{b}\in\mathbb{R}^d, W \in \mathbb{R}^{\mathcal{T}\times d}, \textbf{v} \in \mathbb{R}^{\mathcal{T}}\right\} $$ where the spectral norm of $W$ and $\textbf{v}$ is bounded by $O(1)$, the Frobenius norm of $W$ is bounded from its initialization by $R > 0$, and $σ$ is a Lipschitz activation function. We prove that if $σ$ is element-wise, then the sample complexity of $\mathcal{H}$ has only logarithmic dependency in width and that this complexity is tight, up to logarithmic factors. We further show that the element-wise property of $σ$ is essential for a logarithmic dependency bound in width, in the sense that there exist non-element-wise activation functions whose sample complexity is linear in width, for widths that can be up to exponential in the input dimension. For the upper bound, we use the recent approach for norm-based bounds named Approximate Description Length (ADL) by arXiv:1910.05697. We further develop new techniques and tools for this approach that will hopefully inspire future works.
format Preprint
id arxiv_https___arxiv_org_abs_2211_09634
institution arXiv
publishDate 2022
record_format arxiv
spellingShingle On the Sample Complexity of Two-Layer Networks: Lipschitz vs. Element-Wise Lipschitz Activation
Daniely, Amit
Granot, Elad
Machine Learning
We investigate the sample complexity of bounded two-layer neural networks using different activation functions. In particular, we consider the class $$ \mathcal{H} = \left\{\textbf{x}\mapsto \langle \textbf{v}, σ\circ W\textbf{b} + \textbf{b} \rangle : \textbf{b}\in\mathbb{R}^d, W \in \mathbb{R}^{\mathcal{T}\times d}, \textbf{v} \in \mathbb{R}^{\mathcal{T}}\right\} $$ where the spectral norm of $W$ and $\textbf{v}$ is bounded by $O(1)$, the Frobenius norm of $W$ is bounded from its initialization by $R > 0$, and $σ$ is a Lipschitz activation function. We prove that if $σ$ is element-wise, then the sample complexity of $\mathcal{H}$ has only logarithmic dependency in width and that this complexity is tight, up to logarithmic factors. We further show that the element-wise property of $σ$ is essential for a logarithmic dependency bound in width, in the sense that there exist non-element-wise activation functions whose sample complexity is linear in width, for widths that can be up to exponential in the input dimension. For the upper bound, we use the recent approach for norm-based bounds named Approximate Description Length (ADL) by arXiv:1910.05697. We further develop new techniques and tools for this approach that will hopefully inspire future works.
title On the Sample Complexity of Two-Layer Networks: Lipschitz vs. Element-Wise Lipschitz Activation
topic Machine Learning
url https://arxiv.org/abs/2211.09634