Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Weisfeiler and Leman Go Neural: Higher-order Graph Neural Networks

About

In recent years, graph neural networks (GNNs) have emerged as a powerful neural architecture to learn vector representations of nodes and graphs in a supervised, end-to-end fashion. Up to now, GNNs have only been evaluated empirically -- showing promising results. The following work investigates GNNs from a theoretical point of view and relates them to the $1$-dimensional Weisfeiler-Leman graph isomorphism heuristic ($1$-WL). We show that GNNs have the same expressiveness as the $1$-WL in terms of distinguishing non-isomorphic (sub-)graphs. Hence, both algorithms also have the same shortcomings. Based on this, we propose a generalization of GNNs, so-called $k$-dimensional GNNs ($k$-GNNs), which can take higher-order graph structures at multiple scales into account. These higher-order structures play an essential role in the characterization of social networks and molecule graphs. Our experimental evaluation confirms our theoretical findings as well as confirms that higher-order information is useful in the task of graph classification and regression.

Christopher Morris, Martin Ritzert, Matthias Fey, William L. Hamilton, Jan Eric Lenssen, Gaurav Rattan, Martin Grohe• 2018

Related benchmarks

TaskDatasetResultRank
Graph ClassificationPROTEINS
Accuracy75.5
994
Graph ClassificationMUTAG
Accuracy86.1
862
Graph ClassificationNCI1
Accuracy76.2
501
Graph ClassificationIMDB-B
Accuracy74.2
378
Molecular property predictionQM9 (test)
mu0.493
229
Graph ClassificationMUTAG (10-fold cross-validation)
Accuracy86.1
219
Graph ClassificationMutag (test)
Accuracy86.1
217
Graph ClassificationPROTEINS (10-fold cross-validation)
Accuracy75.5
214
Graph ClassificationPTC-MR
Accuracy60.9
197
Graph ClassificationPROTEINS (test)
Accuracy75.5
180
Showing 10 of 49 rows

Other info

Code

Follow for update