A Recipe for Causal Graph Regression: Confounding Effects Revisited
This work addresses a more challenging regression setting in graph learning, which is incremental as it adapts existing causal techniques to a new task.
The paper tackles the overlooked problem of causal graph regression (CGR) in out-of-distribution scenarios by reshaping confounding effect processing from classification to regression, achieving validated efficacy on benchmarks.
Through recognizing causal subgraphs, causal graph learning (CGL) has risen to be a promising approach for improving the generalizability of graph neural networks under out-of-distribution (OOD) scenarios. However, the empirical successes of CGL techniques are mostly exemplified in classification settings, while regression tasks, a more challenging setting in graph learning, are overlooked. We thus devote this work to tackling causal graph regression (CGR); to this end we reshape the processing of confounding effects in existing CGL studies, which mainly deal with classification. Specifically, we reflect on the predictive power of confounders in graph-level regression, and generalize classification-specific causal intervention techniques to regression through a lens of contrastive learning. Extensive experiments on graph OOD benchmarks validate the efficacy of our proposals for CGR. The model implementation and the code are provided on https://github.com/causal-graph/CGR.