Robustness of Graph Classification: failure modes, causes, and noise-resistant loss in Graph Neural Networks
This addresses robustness issues in GNNs for applied problems with noisy labels, representing an incremental improvement.
The paper tackles the problem of Graph Neural Networks (GNNs) being vulnerable to noisy labels in graph classification tasks, and it introduces two training strategies that enhance robustness without harming performance in noise-free settings.
Graph Neural Networks (GNNs) are powerful at solving graph classification tasks, yet applied problems often contain noisy labels. In this work, we study GNN robustness to label noise, demonstrate GNN failure modes when models struggle to generalise on low-order graphs, low label coverage, or when a model is over-parameterized. We establish both empirical and theoretical links between GNN robustness and the reduction of the total Dirichlet Energy of learned node representations, which encapsulates the hypothesized GNN smoothness inductive bias. Finally, we introduce two training strategies to enhance GNN robustness: (1) by incorporating a novel inductive bias in the weight matrices through the removal of negative eigenvalues, connected to Dirichlet Energy minimization; (2) by extending to GNNs a loss penalty that promotes learned smoothness. Importantly, neither approach negatively impacts performance in noise-free settings, supporting our hypothesis that the source of GNNs robustness is their smoothness inductive bias.