Robust Graph Neural Networks via Unbiased Aggregation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Hou, Zhichao, Feng, Ruiqi, Derr, Tyler, Liu, Xiaorui
Format: Preprint
Published: 2023
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913574677381120
author Hou, Zhichao
Feng, Ruiqi
Derr, Tyler
Liu, Xiaorui
author_facet Hou, Zhichao
Feng, Ruiqi
Derr, Tyler
Liu, Xiaorui
contents The adversarial robustness of Graph Neural Networks (GNNs) has been questioned due to the false sense of security uncovered by strong adaptive attacks despite the existence of numerous defenses. In this work, we delve into the robustness analysis of representative robust GNNs and provide a unified robust estimation point of view to understand their robustness and limitations. Our novel analysis of estimation bias motivates the design of a robust and unbiased graph signal estimator. We then develop an efficient Quasi-Newton Iterative Reweighted Least Squares algorithm to solve the estimation problem, which is unfolded as robust unbiased aggregation layers in GNNs with theoretical guarantees. Our comprehensive experiments confirm the strong robustness of our proposed model under various scenarios, and the ablation study provides a deep understanding of its advantages. Our code is available at https://github.com/chris-hzc/RUNG.
format Preprint
id arxiv_https___arxiv_org_abs_2311_14934
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Robust Graph Neural Networks via Unbiased Aggregation
Hou, Zhichao
Feng, Ruiqi
Derr, Tyler
Liu, Xiaorui
Machine Learning
The adversarial robustness of Graph Neural Networks (GNNs) has been questioned due to the false sense of security uncovered by strong adaptive attacks despite the existence of numerous defenses. In this work, we delve into the robustness analysis of representative robust GNNs and provide a unified robust estimation point of view to understand their robustness and limitations. Our novel analysis of estimation bias motivates the design of a robust and unbiased graph signal estimator. We then develop an efficient Quasi-Newton Iterative Reweighted Least Squares algorithm to solve the estimation problem, which is unfolded as robust unbiased aggregation layers in GNNs with theoretical guarantees. Our comprehensive experiments confirm the strong robustness of our proposed model under various scenarios, and the ablation study provides a deep understanding of its advantages. Our code is available at https://github.com/chris-hzc/RUNG.
title Robust Graph Neural Networks via Unbiased Aggregation
topic Machine Learning
url https://arxiv.org/abs/2311.14934