Investigating Transfer Learning in Graph Neural Networks
This addresses the lack of research on transferability in GNNs, which is important for improving training efficiency and performance in graph-based applications, though it is incremental as it builds on existing GNN methods.
The paper tackles the problem of transfer learning in graph neural networks (GNNs), demonstrating that it is effective and showing that GNNs with inductive operations yield statistically significant improvements in transfer, with results based on experiments using real-world and synthetic data.
Graph neural networks (GNNs) build on the success of deep learning models by extending them for use in graph spaces. Transfer learning has proven extremely successful for traditional deep learning problems: resulting in faster training and improved performance. Despite the increasing interest in GNNs and their use cases, there is little research on their transferability. This research demonstrates that transfer learning is effective with GNNs, and describes how source tasks and the choice of GNN impact the ability to learn generalisable knowledge. We perform experiments using real-world and synthetic data within the contexts of node classification and graph classification. To this end, we also provide a general methodology for transfer learning experimentation and present a novel algorithm for generating synthetic graph classification tasks. We compare the performance of GCN, GraphSAGE and GIN across both the synthetic and real-world datasets. Our results demonstrate empirically that GNNs with inductive operations yield statistically significantly improved transfer. Further we show that similarity in community structure between source and target tasks support statistically significant improvements in transfer over and above the use of only the node attributes.