A Theoretical View of Linear Backpropagation and Its Convergence

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Li, Ziang, Guo, Yiwen, Liu, Haodi, Zhang, Changshui
Format: Preprint
Published: 2021
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911752648654848
author Li, Ziang
Guo, Yiwen
Liu, Haodi
Zhang, Changshui
author_facet Li, Ziang
Guo, Yiwen
Liu, Haodi
Zhang, Changshui
contents Backpropagation (BP) is widely used for calculating gradients in deep neural networks (DNNs). Applied often along with stochastic gradient descent (SGD) or its variants, BP is considered as a de-facto choice in a variety of machine learning tasks including DNN training and adversarial attack/defense. Recently, a linear variant of BP named LinBP was introduced for generating more transferable adversarial examples for performing black-box attacks, by Guo et al. Although it has been shown empirically effective in black-box attacks, theoretical studies and convergence analyses of such a method is lacking. This paper serves as a complement and somewhat an extension to Guo et al.'s paper, by providing theoretical analyses on LinBP in neural-network-involved learning tasks, including adversarial attack and model training. We demonstrate that, somewhat surprisingly, LinBP can lead to faster convergence in these tasks in the same hyper-parameter settings, compared to BP. We confirm our theoretical results with extensive experiments.
format Preprint
id arxiv_https___arxiv_org_abs_2112_11018
institution arXiv
publishDate 2021
record_format arxiv
spellingShingle A Theoretical View of Linear Backpropagation and Its Convergence
Li, Ziang
Guo, Yiwen
Liu, Haodi
Zhang, Changshui
Machine Learning
Cryptography and Security
Computer Vision and Pattern Recognition
Neural and Evolutionary Computing
Backpropagation (BP) is widely used for calculating gradients in deep neural networks (DNNs). Applied often along with stochastic gradient descent (SGD) or its variants, BP is considered as a de-facto choice in a variety of machine learning tasks including DNN training and adversarial attack/defense. Recently, a linear variant of BP named LinBP was introduced for generating more transferable adversarial examples for performing black-box attacks, by Guo et al. Although it has been shown empirically effective in black-box attacks, theoretical studies and convergence analyses of such a method is lacking. This paper serves as a complement and somewhat an extension to Guo et al.'s paper, by providing theoretical analyses on LinBP in neural-network-involved learning tasks, including adversarial attack and model training. We demonstrate that, somewhat surprisingly, LinBP can lead to faster convergence in these tasks in the same hyper-parameter settings, compared to BP. We confirm our theoretical results with extensive experiments.
title A Theoretical View of Linear Backpropagation and Its Convergence
topic Machine Learning
Cryptography and Security
Computer Vision and Pattern Recognition
Neural and Evolutionary Computing
url https://arxiv.org/abs/2112.11018