Test-Time Training for Graph Neural Networks
This addresses generalization issues in graph neural networks for graph classification, but it is incremental as it adapts existing test-time training concepts to GNNs.
The paper tackles the performance gap between training and test sets in graph classification by introducing the first test-time training framework for GNNs, which uses self-supervised learning to adjust models per test sample and shows effectiveness, especially under distribution shifts.
Graph Neural Networks (GNNs) have made tremendous progress in the graph classification task. However, a performance gap between the training set and the test set has often been noticed. To bridge such gap, in this work we introduce the first test-time training framework for GNNs to enhance the model generalization capacity for the graph classification task. In particular, we design a novel test-time training strategy with self-supervised learning to adjust the GNN model for each test graph sample. Experiments on the benchmark datasets have demonstrated the effectiveness of the proposed framework, especially when there are distribution shifts between training set and test set. We have also conducted exploratory studies and theoretical analysis to gain deeper understandings on the rationality of the design of the proposed graph test time training framework (GT3).