Causal invariant geographic network representations with feature and structural distribution shifts
Yuhan Wang, Silu He, Qinyao Luo, Hongyuan Yuan, Ling Zhao, Jiawei Zhu, Haifeng Li
TL;DR
This work tackles the problem of poor OOD generalization in geographic networks caused by feature and structural distribution shifts. It introduces FSM-IRL, a hybrid framework that combines a causal attention-based sampling mechanism for structural shifts with an HSIC-based sample reweighting strategy to decorrelate representation dimensions, thereby emphasizing invariant representations $C$ and suppressing background $S$ dependencies. The method leverages backdoor-adjusted causal estimates and attention-based selection to focus on neighbours with strong causal relevance to labels, while enforcing dimension-wise independence among node representations to reduce spurious correlations. Empirical results across geographic and social datasets show FSM-IRL achieves substantial gains over baselines under various bias settings, demonstrating improved robustness and practical potential for real-world geographic network analysis.
Abstract
The existing methods learn geographic network representations through deep graph neural networks (GNNs) based on the i.i.d. assumption. However, the spatial heterogeneity and temporal dynamics of geographic data make the out-of-distribution (OOD) generalisation problem particularly salient. The latter are particularly sensitive to distribution shifts (feature and structural shifts) between testing and training data and are the main causes of the OOD generalisation problem. Spurious correlations are present between invariant and background representations due to selection biases and environmental effects, resulting in the model extremes being more likely to learn background representations. The existing approaches focus on background representation changes that are determined by shifts in the feature distributions of nodes in the training and test data while ignoring changes in the proportional distributions of heterogeneous and homogeneous neighbour nodes, which we refer to as structural distribution shifts. We propose a feature-structure mixed invariant representation learning (FSM-IRL) model that accounts for both feature distribution shifts and structural distribution shifts. To address structural distribution shifts, we introduce a sampling method based on causal attention, encouraging the model to identify nodes possessing strong causal relationships with labels or nodes that are more similar to the target node. Inspired by the Hilbert-Schmidt independence criterion, we implement a reweighting strategy to maximise the orthogonality of the node representations, thereby mitigating the spurious correlations among the node representations and suppressing the learning of background representations. Our experiments demonstrate that FSM-IRL exhibits strong learning capabilities on both geographic and social network datasets in OOD scenarios.