Principal Neighbourhood Aggregation for Graph Nets

Graph Neural Networks (GNNs) have been shown to be effective models for different predictive tasks on graph-structured data. Recent work on their expressive power has focused on isomorphism tasks and countable feature spaces. We extend this theoretical framework to include continuous features - which occur regularly in real-world input domains and within the hidden layers of GNNs - and we demonstrate the requirement for multiple aggregation functions in this context. Accordingly, we propose Principal Neighbourhood Aggregation (PNA), a novel architecture combining multiple aggregators with degree-scalers (which generalize the sum aggregator). Finally, we compare the capacity of different models to capture and exploit the graph structure via a novel benchmark containing multiple tasks taken from classical graph theory, alongside existing benchmarks from real-world domains, all of which demonstrate the strength of our model. With this work, we hope to steer some of the GNN research towards new aggregation methods which we believe are essential in the search for powerful and robust models.

PDF Abstract NeurIPS 2020 PDF NeurIPS 2020 Abstract
Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Graph Classification CIFAR10 100k PNA Accuracy (%) 70.47 # 4
Graph Property Prediction ogbg-molhiv PNA Test ROC-AUC 0.7905 ± 0.0132 # 21
Validation ROC-AUC 0.8519 ± 0.0099 # 2
Number of params 326081 # 24
Ext. data No # 1
Graph Property Prediction ogbg-molpcba PNA Test AP 0.2838 ± 0.0035 # 18
Validation AP 0.2926 ± 0.0026 # 17
Number of params 6550839 # 8
Ext. data No # 1
Graph Regression ZINC PNA MAE 0.142 # 7

Results from Other Papers


Task Dataset Model Metric Name Metric Value Rank Source Paper Compare
Node Classification PATTERN 100k PNA Accuracy (%) 86.567 # 4

Methods