Salvato in:
Dettagli Bibliografici
Autori principali: Huang, Zheng, Yang, Qihui, Zhou, Dawei, Yan, Yujun
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:https://arxiv.org/abs/2406.04601
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866911914067492864
author Huang, Zheng
Yang, Qihui
Zhou, Dawei
Yan, Yujun
author_facet Huang, Zheng
Yang, Qihui
Zhou, Dawei
Yan, Yujun
contents Although most graph neural networks (GNNs) can operate on graphs of any size, their classification performance often declines on graphs larger than those encountered during training. Existing methods insufficiently address the removal of size information from graph representations, resulting in sub-optimal performance and reliance on backbone models. In response, we propose DISGEN, a novel and model-agnostic framework designed to disentangle size factors from graph representations. DISGEN employs size- and task-invariant augmentations and introduces a decoupling loss that minimizes shared information in hidden representations, with theoretical guarantees for its effectiveness. Our empirical results show that DISGEN outperforms the state-of-the-art models by up to 6% on real-world datasets, underscoring its effectiveness in enhancing the size generalizability of GNNs. Our codes are available at: https://github.com/GraphmindDartmouth/DISGEN.
format Preprint
id arxiv_https___arxiv_org_abs_2406_04601
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Enhancing Size Generalization in Graph Neural Networks through Disentangled Representation Learning
Huang, Zheng
Yang, Qihui
Zhou, Dawei
Yan, Yujun
Machine Learning
Although most graph neural networks (GNNs) can operate on graphs of any size, their classification performance often declines on graphs larger than those encountered during training. Existing methods insufficiently address the removal of size information from graph representations, resulting in sub-optimal performance and reliance on backbone models. In response, we propose DISGEN, a novel and model-agnostic framework designed to disentangle size factors from graph representations. DISGEN employs size- and task-invariant augmentations and introduces a decoupling loss that minimizes shared information in hidden representations, with theoretical guarantees for its effectiveness. Our empirical results show that DISGEN outperforms the state-of-the-art models by up to 6% on real-world datasets, underscoring its effectiveness in enhancing the size generalizability of GNNs. Our codes are available at: https://github.com/GraphmindDartmouth/DISGEN.
title Enhancing Size Generalization in Graph Neural Networks through Disentangled Representation Learning
topic Machine Learning
url https://arxiv.org/abs/2406.04601