Discovering Invariant Neighborhood Patterns for Heterophilic Graphs

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Yang, Jinluan, Zhang, Ruihao, Chen, Zhengyu, Xiao, Teng, Wang, Yueyang, Wu, Fei, Kuang, Kun
Format: Preprint
Veröffentlicht: 2024
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866908490259234816
author Yang, Jinluan
Zhang, Ruihao
Chen, Zhengyu
Xiao, Teng
Wang, Yueyang
Wu, Fei
Kuang, Kun
author_facet Yang, Jinluan
Zhang, Ruihao
Chen, Zhengyu
Xiao, Teng
Wang, Yueyang
Wu, Fei
Kuang, Kun
contents This paper studies the problem of distribution shifts on non-homophilous graphs Mosting existing graph neural network methods rely on the homophilous assumption that nodes from the same class are more likely to be linked. However, such assumptions of homophily do not always hold in real-world graphs, which leads to more complex distribution shifts unaccounted for in previous methods. The distribution shifts of neighborhood patterns are much more diverse on non-homophilous graphs. We propose a novel Invariant Neighborhood Pattern Learning (INPL) to alleviate the distribution shifts problem on non-homophilous graphs. Specifically, we propose the Adaptive Neighborhood Propagation (ANP) module to capture the adaptive neighborhood information, which could alleviate the neighborhood pattern distribution shifts problem on non-homophilous graphs. We propose Invariant Non-Homophilous Graph Learning (INHGL) module to constrain the ANP and learn invariant graph representation on non-homophilous graphs. Extensive experimental results on real-world non-homophilous graphs show that INPL could achieve state-of-the-art performance for learning on large non-homophilous graphs.
format Preprint
id arxiv_https___arxiv_org_abs_2403_10572
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Discovering Invariant Neighborhood Patterns for Heterophilic Graphs
Yang, Jinluan
Zhang, Ruihao
Chen, Zhengyu
Xiao, Teng
Wang, Yueyang
Wu, Fei
Kuang, Kun
Machine Learning
Social and Information Networks
This paper studies the problem of distribution shifts on non-homophilous graphs Mosting existing graph neural network methods rely on the homophilous assumption that nodes from the same class are more likely to be linked. However, such assumptions of homophily do not always hold in real-world graphs, which leads to more complex distribution shifts unaccounted for in previous methods. The distribution shifts of neighborhood patterns are much more diverse on non-homophilous graphs. We propose a novel Invariant Neighborhood Pattern Learning (INPL) to alleviate the distribution shifts problem on non-homophilous graphs. Specifically, we propose the Adaptive Neighborhood Propagation (ANP) module to capture the adaptive neighborhood information, which could alleviate the neighborhood pattern distribution shifts problem on non-homophilous graphs. We propose Invariant Non-Homophilous Graph Learning (INHGL) module to constrain the ANP and learn invariant graph representation on non-homophilous graphs. Extensive experimental results on real-world non-homophilous graphs show that INPL could achieve state-of-the-art performance for learning on large non-homophilous graphs.
title Discovering Invariant Neighborhood Patterns for Heterophilic Graphs
topic Machine Learning
Social and Information Networks
url https://arxiv.org/abs/2403.10572