How Robust Are Graph Neural Networks to Structural Noise?
This addresses robustness issues for users of GNNs in graph-structured data applications, but it is incremental as it builds on existing models and focuses on a specific type of noise.
The paper tackles the problem of robustness in graph neural networks (GNNs) to structural noise, showing that while a representative GNN achieves near-perfect accuracy on node structural identity predictions, it is not robust to added noise, but graph-augmented training can significantly improve robustness under the right conditions.
Graph neural networks (GNNs) are an emerging model for learning graph embeddings and making predictions on graph structured data. However, robustness of graph neural networks is not yet well-understood. In this work, we focus on node structural identity predictions, where a representative GNN model is able to achieve near-perfect accuracy. We also show that the same GNN model is not robust to addition of structural noise, through a controlled dataset and set of experiments. Finally, we show that under the right conditions, graph-augmented training is capable of significantly improving robustness to structural noise.