The DOI matches its Crossref record; 8 of the 8 references it cites were found in a registry or an index.
Stability Properties of Graph Neural Networks
Open accessAbstract
Graph neural networks (GNNs) have emerged as a powerful tool for nonlinear processing of graph signals, exhibiting success in recommender systems, power outage prediction, and motion planning, among others. GNNs consist of a cascade of layers, each of which applies a graph convolution, followed by a pointwise nonlinearity. In this work, we study the impact that changes in the underlying topology have on the output of the GNN. First, we show that GNNs are permutation equivariant, which implies that they effectively exploit internal symmetries of the underlying topology. Then, we prove that graph convolutions with integral Lipschitz filters, in combination with the frequency mixing effect of the corresponding nonlinearities, yields an architecture that is both stable to small changes in the underlying topology, and discriminative of information located at high frequencies. These are two properties that cannot simultaneously hold when using only linear graph filters, which are either discriminative or stable, thus explaining the superior performance of GNNs.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
LeCun, Y., Bengio, Y., & Hinton, G. (2015) · Nature · doi:10.1038/nature14539 · Cited by 84,126
Pan, S. J., & Yang, Q. (2010) · IEEE Transactions on Knowledge and Data Engineering · doi:10.1109/tkde.2009.191 · Cited by 23,387 · cited 1×
“ample, several works have studied, experimentally, the concept of transfer learning, which considers the ability to train a CNN on one dataset, and then sucessfully use it on another, similar dataset [10]. In the theoretical realm, [11] opted for controlling for one of the sources of uncertainty, by fixing the bank of…”
BackgroundDaubechies, I. (1992) · doi:10.1137/1.9781611970104 · Cited by 15,979
Shuman, D. I., Narang, S. K., Frossard, P., Ortega, A., & Vandergheynst, P. (2013) · IEEE Signal Processing Magazine · doi:10.1109/msp.2012.2235192 · Cited by 4,421 · cited 1×
“The matrix representation S is extended here to be a graph shift operator (GSO) [5], [7] which we formally define as a matrix S that respects the sparsity of the graph, namely, s ij = [ S ] ij = 0 whenever i = j and ( j, i ) / ∈ E .”
BackgroundGraph Convolutional Neural Networks for Web-Scale Recommender Systems
Ying, R., He, R., Chen, K., Eksombatchai, P., Hamilton, W. L., & Leskovec, J. (2018) · Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining · doi:10.1145/3219819.3219890 · Cited by 3,733
Discrete Signal Processing on Graphs
Sandryhaila, A., & Moura, J. M. F. (2013) · IEEE Transactions on Signal Processing · doi:10.1109/tsp.2013.2238935 · Cited by 1,531 · cited 2×
“xhibit a regular inherent structure that can be effectively exploited by convolutions. Data elements are, instead, related by arbitrary pairwise relationships described by an underlying graph support [16]–[19]. Graph neural networks (GNNs) have emerged arXiv:1905.04497v1 [cs.LG] 11 May 2019 2 as successful architectur…”
BackgroundIshaque, K., Salam, Z., Amjad, M., & Mekhilef, S. (2012) · IEEE Transactions on Power Electronics · doi:10.1109/tpel.2012.2185713 · Cited by 1,100
Chen, T., Wang, Z., Li, L., Li, D., Li, Z., Chang, X., Bian, P., Liang, G., Wang, Q., & Xie, T. (2025) · Schloss Dagstuhl – Leibniz-Zentrum für Informatik · doi:10.4230/lipics.ecoop.2025.6
Cited by
A Realization-Theoretic Foundation for Fault Diagnosis, with Diagnosability Equivalence Predictions
Bronstein, M. M., Bruna, J., Cohen, T., & Veličković, P. (2021) · arXiv (Cornell University) · doi:10.4230/oasics.dx.2026.1 · Cited by 519
Self-citationGraphLIME: Local Interpretable Model Explanations for Graph Neural Networks
Huang, Q., Yamada, M., Tian, Y., Singh, D., & Chang, Y. (2023) · IEEE Transactions on Knowledge and Data Engineering · doi:10.1109/tkde.2022.3187455 · Cited by 424
Kernel Mean Embedding of Distributions: A Review and Beyond
Muandet, K., Fukumizu, K., Sriperumbudur, B., & Schölkopf, B. (2017) · Foundations and Trends® in Machine Learning · doi:10.1561/2200000060 · Cited by 409
She, C., Sun, C., Gu, Z., Li, Y., Yang, C., Poor, H. V., & Vucetic, B. (2021) · Proceedings of the IEEE · doi:10.1109/jproc.2021.3053601 · Cited by 405
Optimal Wireless Resource Allocation With Random Edge Graph Neural Networks
Eisen, M., & Ribeiro, A. (2020) · IEEE Transactions on Signal Processing · doi:10.1109/tsp.2020.2988255 · Cited by 327
Self-citation
Metadata
- Authors
- Gama, F., Bruna, J., & Ribeiro, A.
- Venue
- IEEE Transactions on Signal Processing
- Published
- 2020
- Volume
- 68
- Pages
- 5680-5695
- ISSN
- 1053-587X (ISSN-L)
- 1941-0476
- OpenAlex
- W2943959761
- Cited by
- 204
- References
- 8
- FWCI
- 11.65
- Type
- journal article
- Language
- en
- Open access
- green
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.