Graph Mixup with Soft Alignments
This addresses a domain-specific problem for researchers and practitioners in graph machine learning by providing an incremental enhancement to existing mixup techniques for graph data.
The paper tackled the challenge of applying mixup data augmentation to graph data, which lacks node-level correspondence, by proposing S-Mixup, a method using soft alignments to enable direct mixing of graph pairs, resulting in improved performance and generalization of graph neural networks on classification tasks and increased robustness against noisy labels.
We study graph data augmentation by mixup, which has been used successfully on images. A key operation of mixup is to compute a convex combination of a pair of inputs. This operation is straightforward for grid-like data, such as images, but challenging for graph data. The key difficulty lies in the fact that different graphs typically have different numbers of nodes, and thus there lacks a node-level correspondence between graphs. In this work, we propose S-Mixup, a simple yet effective mixup method for graph classification by soft alignments. Specifically, given a pair of graphs, we explicitly obtain node-level correspondence via computing a soft assignment matrix to match the nodes between two graphs. Based on the soft assignments, we transform the adjacency and node feature matrices of one graph, so that the transformed graph is aligned with the other graph. In this way, any pair of graphs can be mixed directly to generate an augmented graph. We conduct systematic experiments to show that S-Mixup can improve the performance and generalization of graph neural networks (GNNs) on various graph classification tasks. In addition, we show that S-Mixup can increase the robustness of GNNs against noisy labels.