Enhancing Size Generalization in Graph Neural Networks through Disentangled Representation Learning
This addresses the size generalization issue in GNNs for applications like social networks or molecular analysis, representing a novel method for a known bottleneck.
The paper tackles the problem of graph neural networks (GNNs) performing poorly on graphs larger than those seen during training by proposing DISGEN, a framework that disentangles size factors from graph representations, resulting in up to 6% improvement over state-of-the-art models on real-world datasets.
Although most graph neural networks (GNNs) can operate on graphs of any size, their classification performance often declines on graphs larger than those encountered during training. Existing methods insufficiently address the removal of size information from graph representations, resulting in sub-optimal performance and reliance on backbone models. In response, we propose DISGEN, a novel and model-agnostic framework designed to disentangle size factors from graph representations. DISGEN employs size- and task-invariant augmentations and introduces a decoupling loss that minimizes shared information in hidden representations, with theoretical guarantees for its effectiveness. Our empirical results show that DISGEN outperforms the state-of-the-art models by up to 6% on real-world datasets, underscoring its effectiveness in enhancing the size generalizability of GNNs. Our codes are available at: https://github.com/GraphmindDartmouth/DISGEN.