Structural Alignment Improves Graph Test-Time Adaptation

TL;DR

This paper introduces Test-Time Structural Alignment (TSA), a novel method for graph test-time adaptation that aligns graph structures during inference without retraining, improving performance under distribution shifts.

Contribution

The paper proposes TSA, a new algorithm for graph test-time adaptation that leverages structural alignment strategies to enhance model robustness without retraining.

Findings

TSA outperforms existing methods on synthetic and real datasets.

TSA effectively handles distribution shifts in graph data.

Theoretical analysis supports TSA's strategies for structural alignment.

Abstract

Graph-based learning excels at capturing interaction patterns in diverse domains like recommendation, fraud detection, and particle physics. However, its performance often degrades under distribution shifts, especially those altering network connectivity. Current methods to address these shifts typically require retraining with the source dataset, which is often infeasible due to computational or privacy limitations. We introduce Test-Time Structural Alignment (TSA), a novel algorithm for Graph Test-Time Adaptation (GTTA) that adapts a pretrained model to align graph structures during inference without the cost of retraining. Grounded in a theoretical understanding of graph data distribution shifts, TSA employs three synergistic strategies: uncertainty-aware neighborhood weighting to accommodate neighbor label distribution shifts, adaptive balancing of self-node and aggregated neighborhood representations based on their signal-to-noise ratio, and decision boundary refinement to correct residual label and feature shifts. Extensive experiments on synthetic and real-world datasets demonstrate TSA's consistent outperformance of both non-graph TTA methods and state-of-the-art GTTA baselines.