Timezone: »

Hierarchical Graph Representation Learning with Differentiable Pooling
Zhitao Ying · Jiaxuan You · Christopher Morris · Xiang Ren · Will Hamilton · Jure Leskovec

Tue Dec 04 02:00 PM -- 04:00 PM (PST) @ Room 210 #14

Recently, graph neural networks (GNNs) have revolutionized the field of graph representation learning through effectively learned node embeddings, and achieved state-of-the-art results in tasks such as node classification and link prediction. However, current GNN methods are inherently flat and do not learn hierarchical representations of graphs---a limitation that is especially problematic for the task of graph classification, where the goal is to predict the label associated with an entire graph. Here we propose DiffPool, a differentiable graph pooling module that can generate hierarchical representations of graphs and can be combined with various graph neural network architectures in an end-to-end fashion. DiffPool learns a differentiable soft cluster assignment for nodes at each layer of a deep GNN, mapping nodes to a set of clusters, which then form the coarsened input for the next GNN layer. Our experimental results show that combining existing GNN methods with DiffPool yields an average improvement of 5-10% accuracy on graph classification benchmarks, compared to all existing pooling approaches, achieving a new state-of-the-art on four out of five benchmark datasets.

Author Information

Zhitao Ying (Stanford University)
Jiaxuan You (Stanford University)

2nd CS PhD student in Stanford

Christopher Morris (TU Dortmund University)
Xiang Ren (University of Southern California)
Will Hamilton (McGill University / FAIR)
Jure Leskovec (Stanford University and Pinterest)

Related Events (a corresponding poster, oral, or spotlight)

More from the Same Authors