GATTA：结合测试时数据增强的图主动学习

中文摘要 AI 辅助

测试时数据增强（Test-time augmentation，TTA）已被证明可有效提升计算机视觉领域模型的鲁棒性与不确定性估计能力，但其在图结构化数据中的应用仍未得到充分探索。我们提出GATTA（Graph Active Learning with Test-Time Augmentation，结合测试时数据增强的图主动学习）框架，该框架通过聚合多个增强视图的预测结果来生成更可靠的不确定性估计，以此增强主动学习性能。为解决保持标签一致性的图增强难题，GATTA引入了基于一致性的过滤机制，用于丢弃预测结果不可靠的增强视图。我们在多个图数据集、图神经网络（GNN）架构及获取策略上对GATTA进行了系统评估。结果表明，基于简单不确定性的方法（如熵（Entropy）、最小置信度（Least Confidence））从TTA中获益最多，其性能可与更复杂、计算成本更高的方法相媲美。GATTA可跨架构泛化，且性能优于模型侧集成方法（如MC Dropout）。我们进一步证明，GATTA可随集成规模与图规模高效扩展。对增强类型、强度及过滤策略的大量分析为有效部署提供了实用指南。研究发现，为简单方法结合TTA是实现高性能主动学习的更高效途径，相较于设计复杂的获取函数，该方法能让从业者以更低的计算开销和更简化的实现复杂度获得具有竞争力的结果。

英文摘要

Test-time augmentation (TTA) has proven effective for improving model robustness and uncertainty estimation in computer vision, yet its application to graph-structured data remains largely unexplored. We introduce GATTA (Graph Active Learning with Test-Time Augmentation), a framework for enhancing active learning by aggregating predictions across multiple augmented views to produce more reliable uncertainty estimates. To address the challenge of label-preserving graph augmentations, GATTA incorporates a consistency-based filtering mechanism that discards augmented views yielding unreliable predictions. We systematically evaluate GATTA across multiple graph datasets, GNN architectures, and acquisition strategies. Our results show that simple uncertainty-based methods, such as Entropy and Least Confidence, benefit most from TTA, achieving performance competitive with more sophisticated and computationally expensive approaches. GATTA generalizes across architectures, outperforms model-side ensemble methods such as MC Dropout. We further show that GATTA scales efficiently with both ensemble size and graph size. Extensive analysis of augmentation types, strengths, and filtering strategies provides practical guidelines for effective deployment. Our findings demonstrate that augmenting simple methods with TTA offers a more efficient path to strong active learning performance than engineering complex acquisition functions, enabling practitioners to achieve competitive results with lower computational overhead and reduced implementation complexity.

发表机构

• Budapest University of Technology and Economics（布达佩斯技术与经济大学）
• KU Leuven（鲁汶大学）

机构由 AI 辅助整理，请以论文原文为准。