Uncertainty Propagation in Node Classification

被引:0
|
作者
Xu, Zhao [1 ]
Lawrence, Carolin [1 ]
Shaker, Ammar [1 ]
Siarheyeu, Raman [1 ]
机构
[1] NEC Labs Europe, Heidelberg, Germany
关键词
Uncertainty propagation; Bayesian modeling; Graph neural networks;
D O I
10.1109/ICDM54844.2022.00167
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Quantifying predictive uncertainty of neural networks has recently attracted increasing attention. In this work, we focus on measuring uncertainty of graph neural networks (GNNs) for the task of node classification. Most existing GNNs model message passing among nodes. The messages are often deterministic. Questions naturally arise: Does there exist uncertainty in the messages? How could we propagate such uncertainty over a graph together with messages? To address these issues, we propose a Bayesian uncertainty propagation (BUP) method, which embeds GNNs in a Bayesian modeling framework, and models predictive uncertainty of node classification with Bayesian confidence of predictive probability and uncertainty of messages. Our method proposes a novel uncertainty propagation mechanism inspired by Gaussian models. Moreover, we present an uncertainty oriented loss for node classification that allows the GNNs to clearly integrate predictive uncertainty in learning procedure. Consequently, the training examples with large predictive uncertainty will be penalized. We demonstrate the BUP with respect to prediction reliability and out-of-distribution (OOD) predictions. The learned uncertainty is also analyzed in depth. The relations between uncertainty and graph topology, as well as predictive uncertainty in the OOD cases are investigated with extensive experiments. The empirical results with popular benchmark datasets demonstrate the superior performance of the proposed method.
引用
收藏
页码:1275 / 1280
页数:6
相关论文
共 50 条
  • [1] Chain of Propagation Prompting for Node Classification
    Zhu, Yonghua
    Deng, Zhenyun
    Chen, Yang
    Amor, Robert
    Witbrock, Michael
    PROCEEDINGS OF THE 31ST ACM INTERNATIONAL CONFERENCE ON MULTIMEDIA, MM 2023, 2023, : 3012 - 3020
  • [2] Particle Propagation Model for Dynamic Node Classification
    Li, Wenzheng
    Yin, Dechun
    Yuan, Deyu
    Wang, Benyu
    Gu, Yijun
    IEEE ACCESS, 2020, 8 : 140205 - 140215
  • [3] Exploring Node Classification Uncertainty in Graph Neural Networks
    Islam, Md. Farhadul
    Zabeen, Sarah
    Bin Rahman, Fardin
    Islam, Md. Azharul
    Bin Kibria, Fahmid
    Manab, Meem Arafat
    Karim, Dewan Ziaul
    Rasel, Annajiat Alim
    PROCEEDINGS OF THE 2023 ACM SOUTHEAST CONFERENCE, ACMSE 2023, 2023, : 186 - 190
  • [4] GraphHop: An Enhanced Label Propagation Method for Node Classification
    Xie, Tian
    Wang, Bin
    Kuo, C-C Jay
    IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS, 2023, 34 (11) : 9287 - 9301
  • [5] Label Efficient Regularization and Propagation for Graph Node Classification
    Xie, Tian
    Kannan, Rajgopal
    Kuo, C. -C. Jay
    IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2023, 45 (12) : 14856 - 14871
  • [6] Graph Posterior Network: Bayesian Predictive Uncertainty for Node Classification
    Stadler, Maximilian
    Charpentier, Bertrand
    Geisler, Simon
    Zuegner, Daniel
    Guennemann, Stephan
    ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 34 (NEURIPS 2021), 2021, 34
  • [7] Node classification using kernel propagation in graph neural networks
    Prakash, Sakthi Kumar Arul
    Tucker, Conrad S.
    EXPERT SYSTEMS WITH APPLICATIONS, 2021, 174
  • [8] Adaptive Node Embedding Propagation for Semi-supervised Classification
    Ogawa, Yuya
    Maekawa, Seiji
    Sasaki, Yuya
    Fujiwara, Yasuhiro
    Onizuka, Makoto
    MACHINE LEARNING AND KNOWLEDGE DISCOVERY IN DATABASES, ECML PKDD 2021: RESEARCH TRACK, PT II, 2021, 12976 : 417 - 433
  • [9] Propagation with Adaptive Mask then Training for Node Classification on Attributed Networks
    Chen, Jinsong
    Li, Boyu
    He, Qiuting
    He, Kun
    SSRN, 2022,
  • [10] Piecewise point classification for uncertainty propagation with nonlinear limit states
    Wang, Zequn
    STRUCTURAL AND MULTIDISCIPLINARY OPTIMIZATION, 2017, 56 (02) : 285 - 296