Distribution Preserving Graph Representation Learning
This addresses generalization issues in graph representation learning, which is important for applications like social network analysis or bioinformatics, but it appears incremental as it builds on existing expressive GNN backbones.
The paper tackles the problem of graph neural networks (GNNs) overfitting to training data and losing generalization ability by proposing DP-GNN, a framework that preserves distribution information in representations, and it achieves state-of-the-art performance on graph classification benchmarks.
Graph neural network (GNN) is effective to model graphs for distributed representations of nodes and an entire graph. Recently, research on the expressive power of GNN attracted growing attention. A highly-expressive GNN has the ability to generate discriminative graph representations. However, in the end-to-end training process for a certain graph learning task, a highly-expressive GNN risks generating graph representations overfitting the training data for the target task, while losing information important for the model generalization. In this paper, we propose Distribution Preserving GNN (DP-GNN) - a GNN framework that can improve the generalizability of expressive GNN models by preserving several kinds of distribution information in graph representations and node representations. Besides the generalizability, by applying an expressive GNN backbone, DP-GNN can also have high expressive power. We evaluate the proposed DP-GNN framework on multiple benchmark datasets for graph classification tasks. The experimental results demonstrate that our model achieves state-of-the-art performances.