What functions can Graph Neural Networks compute on random graphs? The role of Positional Encoding

被引:0
|
作者
Keriven, Nicolas [1 ]
Vaiter, Samuel [2 ]
机构
[1] CNRS, IRISA, Rennes, France
[2] CNRS, LJAD, Nice, France
关键词
D O I
暂无
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
We aim to deepen the theoretical understanding of Graph Neural Networks (GNNs) on large graphs, with a focus on their expressive power. Existing analyses relate this notion to the graph isomorphism problem, which is mostly relevant for graphs of small sizes, or studied graph classification or regression tasks, while prediction tasks on nodes are far more relevant on large graphs. Recently, several works showed that, on very general random graphs models, GNNs converge to certains functions as the number of nodes grows. In this paper, we provide a more complete and intuitive description of the function space generated by equivariant GNNs for node-tasks, through general notions of convergence that encompass several previous examples. We emphasize the role of input node features, and study the impact of node Positional Encodings (PEs), a recent line of work that has been shown to yield state-of-the-art results in practice. Through the study of several examples of PEs on large random graphs, we extend previously known universality results to significantly more general models. Our theoretical results hint at some normalization tricks, which is shown numerically to have a positive impact on GNN generalization on synthetic and real data. Our proofs contain new concentration inequalities of independent interest.
引用
收藏
页数:27
相关论文
共 50 条
  • [1] Watermarking Graph Neural Networks by Random Graphs
    Zhao, Xiangyu
    Wu, Hanzhou
    Zhang, Xinpeng
    9TH INTERNATIONAL SYMPOSIUM ON DIGITAL FORENSICS AND SECURITY (ISDFS'21), 2021,
  • [2] On the Universality of Graph Neural Networks on Large Random Graphs
    Keriven, Nicolas
    Bietti, Alberto
    Vaiter, Samuel
    ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 34 (NEURIPS 2021), 2021, 34
  • [3] Graph Random Neural Networks for Semi-Supervised Learning on Graphs
    Feng, Wenzheng
    Zhang, Jie
    Dong, Yuxiao
    Han, Yu
    Luan, Huanbo
    Xu, Qian
    Yang, Qiang
    Kharlamov, Evgeny
    Tang, Jie
    ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 33, NEURIPS 2020, 2020, 33
  • [4] On Positional and Structural Node Features for Graph Neural Networks on Non-attributed Graphs
    Cui, Hejie
    Lu, Zijie
    Li, Pan
    Yang, Carl
    PROCEEDINGS OF THE 31ST ACM INTERNATIONAL CONFERENCE ON INFORMATION AND KNOWLEDGE MANAGEMENT, CIKM 2022, 2022, : 3898 - 3902
  • [5] Graphs, Convolutions, and Neural Networks: From Graph Filters to Graph Neural Networks
    Gama, Fernando
    Isufi, Elvin
    Leus, Geert
    Ribeiro, Alejandro
    IEEE SIGNAL PROCESSING MAGAZINE, 2020, 37 (06) : 128 - 138
  • [6] Information encoding by deep neural networks: what can we learn?
    ten Bosch, L.
    Boves, L.
    19TH ANNUAL CONFERENCE OF THE INTERNATIONAL SPEECH COMMUNICATION ASSOCIATION (INTERSPEECH 2018), VOLS 1-6: SPEECH RESEARCH FOR EMERGING MARKETS IN MULTILINGUAL SOCIETIES, 2018, : 1457 - 1461
  • [7] Positional Encoder Graph Neural Networks for Geographic Data
    Klemmer, Konstantin
    Safir, Nathan
    Neill, Daniel B.
    INTERNATIONAL CONFERENCE ON ARTIFICIAL INTELLIGENCE AND STATISTICS, VOL 206, 2023, 206
  • [8] Random Walk Graph Neural Networks
    Nikolentzos, Giannis
    Vazirgiannis, Michalis
    ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 33, NEURIPS 2020, 2020, 33
  • [9] Cell Tower Localisation using Graph Convolutional Networks and Positional Encoding
    Goel, Sanchit
    Soman, Sumit
    Mangal, Roopesh
    Vuppala, Sunil Kumar
    Banerjee, Jyotirmoy
    PROCEEDINGS OF 7TH JOINT INTERNATIONAL CONFERENCE ON DATA SCIENCE AND MANAGEMENT OF DATA, CODS-COMAD 2024, 2024, : 375 - 383
  • [10] Neural Redshift: Random Networks are not Random Functions
    Teney, Damien
    Nicolicioiu, Armand Mihai
    Hartmann, Valentin
    Abbasnejad, Ehsan
    2024 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION, CVPR 2024, 2024, : 4786 - 4796