Abstract
Recent research in the domain of computed graph embeddings has shown that graph homomorphism numbers constitute expressive features that are well-suited for machine learning tasks such as graph classification. In this work-in-progress paper, we attempt to make this methodology scalable by obtaining additive approximations to graph homomorphism densities via a simple sampling algorithm. We show in experiments that these approximate homomorphism densities perform as well as homomorphism numbers on standard graph classification datasets. Moreover, we show that, unlike algorithms that compute homomorphism numbers, our sampling algorithm is highly scalable to larger graphs.
Keywords
Supported by Agence Nationale de la Recherche (ANR), projects STAP (ANR-17-CE23-0021) and ESIGMA (ANR-17-CE23-0010). F. Yger acknowledges the support of the ANR as part of the “Investissements d’avenir” program (ANR-19-P3IA-0001, PRAIRIE 3IA Institute).
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
Standard datasets contain relatively sparse graphs which means that the corresponding complement graphs are dense. Dense graphs have larger homomorphism densities which are easier to detect at fixed precision and more amenable for training machine learning models.
References
Beaujean, P., Sikora, F., Yger, F.: Scaling up graph homomorphism features with efficient data structures. In: ICLR 2021 Workshop on Geometrical and Topological Representation Learning (2021). https://openreview.net/forum?id=EwT8NpZIth8
Beaujean, P., Sikora, F., Yger, F.: Scaling up graph homomorphism for classification via sampling. CoRR abs/2104.04040 (2021). https://arxiv.org/abs/2104.04040
Bouritsas, G., Frasca, F., Zafeiriou, S., Bronstein, M.M.: Improving graph neural network expressivity via subgraph isomorphism counting. CoRR abs/2006.09252 (2020). https://arxiv.org/abs/2006.09252
Bronstein, M.M., Bruna, J., LeCun, Y., Szlam, A., Vandergheynst, P.: Geometric deep learning: going beyond Euclidean data. IEEE Signal Process. Mag. 34(4), 18–42 (2017)
Bulatov, A., Živný, S.: Approximate counting CSP seen from the other side. ACM Trans. Comput. Theory 12, 1–19 (2020). https://doi.org/10.1145/3389390
Curticapean, R., Dell, H., Marx, D.: Homomorphisms are a good basis for counting small subgraphs. In: Proceedings of the 49th Annual ACM SIGACT Symposium on Theory of Computing, pp. 210–223 (2017)
Dalmau, V., Jonsson, P.: The complexity of counting homomorphisms seen from the other side. Theoret. Comput. Sci. 329(1–3), 315–323 (2004)
Díaz, J., Serna, M., Thilikos, D.M.: Counting H-colorings of partial k-trees. Theor. Comput. Sci. 281(1–2), 291–309 (2002)
Errica, F., Podda, M., Bacciu, D., Micheli, A.: A fair comparison of graph neural networks for graph classification. In: ICLR (2019)
Grohe, M.: word2vec, node2vec, graph2vec, x2vec: towards a theory of vector embeddings of structured data. In: PODS (2020)
Hu, W., Fey, M., Ren, H., Nakata, M., Dong, Y., Leskovec, J.: OGB-LSC: a large-scale challenge for machine learning on graphs. arXiv preprint arXiv:2103.09430 (2021)
Lovász, L.: Operations with structures. Acta Mathematica Academiae Scientiarum Hungarica (1967)
Lovász, L.: Large networks and graph limits. American Mathematical Soc. (2012)
Morris, C., Kriege, N.M., Bause, F., Kersting, K., Mutzel, P., Neumann, M.: TUDataset: A collection of benchmark datasets for learning with graphs. In: ICML Workshop on Graph Representation Learning and Beyond (GRL+) (2020). www.graphlearning.io
NT, H., Maehara, T.: Graph homomorphism convolution. In: ICML (2020)
Shervashidze, N., Vishwanathan, S., Petri, T., Mehlhorn, K., Borgwardt, K.: Efficient graphlet kernels for large graph comparison. In: AIStat (2009)
Wu, Z., Pan, S., Chen, F., Long, G., Zhang, C., Philip, S.Y.: A comprehensive survey on graph neural networks. IEEE Trans. Neural Networks Learn. Syst. (2020)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2021 Springer Nature Switzerland AG
About this paper
Cite this paper
Beaujean, P., Sikora, F., Yger, F. (2021). Graph Homomorphism Features: Why Not Sample?. In: Kamp, M., et al. Machine Learning and Principles and Practice of Knowledge Discovery in Databases. ECML PKDD 2021. Communications in Computer and Information Science, vol 1524. Springer, Cham. https://doi.org/10.1007/978-3-030-93736-2_17
Download citation
DOI: https://doi.org/10.1007/978-3-030-93736-2_17
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-93735-5
Online ISBN: 978-3-030-93736-2
eBook Packages: Computer ScienceComputer Science (R0)