Resist Label Noise with PGM for Graph Neural Networks
Not on the map. Why?
Abstract (arXiv)
While robust graph neural networks (GNNs) have been widely studied for graph perturbation and attack, those for label noise have received significantly less attention. Most existing methods heavily rely on the label smoothness assumption to correct noisy labels, which adversely affects their performance on heterophilous graphs. Further, they generally perform poorly in high noise-rate scenarios. To address these problems, in this paper, we propose a novel probabilistic graphical model (PGM) based framework LNP. Given a noisy label set and a clean label set, our goal is to maximize the likelihood of labels in the clean set. We first present LNP-v1, which generates clean labels based on graphs only in the Bayesian network. To further leverage the information of clean labels in the noisy label set, we put forward LNP-v2, which incorporates the noisy label set into the Bayesian network to generate clean labels. The generative process can then be used to predict labels for unlabeled nodes. We conduct extensive experiments to show the robustness of LNP on varying noise types and rates, and also on graphs with different heterophilies. In particular, we show that LNP can lead to inspiring performance in high noise-rate situations.
Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 29 of its 40 references that are in the map. First posted on arXiv 3 Nov 2023.
Builds on
29 of 40 references.
References · 29
- 2016Semi-Supervised Classification with Graph Convolutional Networkscited by 11,282
- 2017Inductive Representation Learning on Large Graphscited by 6,284
- 2018Co-teaching: Robust Training of Deep Neural Networks with Extremely Noisy Labelscited by 1,053
- 2018Generalized Cross Entropy Loss for Training Deep Neural Networks with Noisy Labelscited by 994
- 2020Geom-GCN: Geometric Graph Convolutional Networkscited by 821
- 2020DivideMix: Learning with Noisy Labels as Semi-supervised Learningcited by 691
- 2016Making Deep Neural Networks Robust to Label Noise: a Loss Correction Approachcited by 663
- 2020Learning from Noisy Labels with Deep Neural Networks: A Surveycited by 580
- 2020Graph Neural Networks in Recommender Systems: A Surveycited by 493
- 2022GraphMAE: Self-Supervised Masked Graph Autoencoderscited by 443
- 2019How does Disagreement Help Generalization against Label Corruption?cited by 390
- 2021Large Scale Learning on Non-Homophilous Graphs: New Benchmarks and Strong Simple Methodscited by 309
17 more
- 2020Combating noisy labels by agreement: A joint training method with co-regularizationcited by 289
- 2020Normalized Loss Functions for Deep Learning with Noisy Labelscited by 257
- 2019A Fair Comparison of Graph Neural Networks for Graph Classificationcited by 244
- 2020GNNGuard: Defending Graph Neural Networks against Adversarial Attackscited by 195
- 2020A Semi-supervised Graph Attentive Network for Financial Fraud Detectioncited by 150
- 2020Self-Adaptive Training: beyond Empirical Risk Minimizationcited by 126
- 2020AM-GCN: Adaptive Multi-channel Graph Convolutional Networkscited by 121
- 2021NRGNN: Learning a Label Noise-Resistant Graph Neural Network on Sparsely and Noisily Labeled Graphscited by 73
- 2019Graph-Revised Convolutional Networkcited by 58
- 2021Open-set Label Noise Can Improve Robustness Against Inherent Label Noisecited by 58
- 2017Deep Learning from Noisy Image Labels with Quality Embeddingcited by 46
- 2020Learning to Extrapolate Knowledge: Transductive Few-shot Out-of-Graph Link Predictioncited by 46
- 2020Weakly Supervised Learning with Side Information for Noisy Labeled Imagescited by 30
- 2019Learning Graph Neural Networks with Noisy Labelscited by 29
- 2022Robust Training of Graph Neural Networks via Noise Governancecited by 29
- 2021Unified Robust Training for Graph NeuralNetworks against Label Noisecited by 21
- 2022On Learning Contrastive Representations for Learning with Noisy Labelscited by 21
Built on by
Not cited yet.