Saved in:
Bibliographic Details
Main Authors: Liu, Songtao, Chen, Jinghui, Fu, Tianfan, Lin, Lu, Zitnik, Marinka, Wu, Dinghao
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2406.02059
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914822077022208
author Liu, Songtao
Chen, Jinghui
Fu, Tianfan
Lin, Lu
Zitnik, Marinka
Wu, Dinghao
author_facet Liu, Songtao
Chen, Jinghui
Fu, Tianfan
Lin, Lu
Zitnik, Marinka
Wu, Dinghao
contents This paper introduces a min-max optimization formulation for the Graph Signal Denoising (GSD) problem. In this formulation, we first maximize the second term of GSD by introducing perturbations to the graph structure based on Laplacian distance and then minimize the overall loss of the GSD. By solving the min-max optimization problem, we derive a new variant of the Graph Diffusion Convolution (GDC) architecture, called Graph Adversarial Diffusion Convolution (GADC). GADC differs from GDC by incorporating an additional term that enhances robustness against adversarial attacks on the graph structure and noise in node features. Moreover, GADC improves the performance of GDC on heterophilic graphs. Extensive experiments demonstrate the effectiveness of GADC across various datasets. Code is available at https://github.com/SongtaoLiu0823/GADC.
format Preprint
id arxiv_https___arxiv_org_abs_2406_02059
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Graph Adversarial Diffusion Convolution
Liu, Songtao
Chen, Jinghui
Fu, Tianfan
Lin, Lu
Zitnik, Marinka
Wu, Dinghao
Machine Learning
This paper introduces a min-max optimization formulation for the Graph Signal Denoising (GSD) problem. In this formulation, we first maximize the second term of GSD by introducing perturbations to the graph structure based on Laplacian distance and then minimize the overall loss of the GSD. By solving the min-max optimization problem, we derive a new variant of the Graph Diffusion Convolution (GDC) architecture, called Graph Adversarial Diffusion Convolution (GADC). GADC differs from GDC by incorporating an additional term that enhances robustness against adversarial attacks on the graph structure and noise in node features. Moreover, GADC improves the performance of GDC on heterophilic graphs. Extensive experiments demonstrate the effectiveness of GADC across various datasets. Code is available at https://github.com/SongtaoLiu0823/GADC.
title Graph Adversarial Diffusion Convolution
topic Machine Learning
url https://arxiv.org/abs/2406.02059