arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2206.01570cs.LG
收藏

用于节点分类的图神经网络的标定

On Calibration of Graph Neural Networks for Node Classification

LMU Munich(慕尼黑大学);Siemens AG(西门子股份公司)

Tong Liu, Yushan Liu, Marcel Hildebrandt, Mitchell Joblin, Hang Li, Volker Tresp

摘要:

图可以通过用节点和边来表示实体及其交互来模拟真实世界的复杂系统。为了更好地利用图结构,已经开发了图神经网络,其学习用于诸如节点分类和链接预测的任务的实体和边缘嵌入。这些模型在准确性方面实现了良好的性能,但与预测相关联的置信度得分可能没有被校准。这意味着分数可能无法反映预测事件的真实概率,这对于安全关键型应用程序尤为重要。尽管图神经网络被用于广泛的任务,但是其校准尚未被充分探索。研究了节点分类的图神经网络的标定问题,研究了现有后处理标定方法的效果,分析了模型容量、图密度和一种新的损失函数对标定的影响。此外,我们提出了一种拓扑感知的校准方法,考虑到相邻节点,并产生改进的校准相比,基线方法。

英文摘要:

Graphs can model real-world, complex systems by representing entities and their interactions in terms of nodes and edges. To better exploit the graph structure, graph neural networks have been developed, which learn entity and edge embeddings for tasks such as node classification and link prediction. These models achieve good performance with respect to accuracy, but the confidence scores associated with the predictions might not be calibrated. That means that the scores might not reflect the ground-truth probabilities of the predicted events, which would be especially important for safety-critical applications. Even though graph neural networks are used for a wide range of tasks, the calibration thereof has not been sufficiently explored yet. We investigate the calibration of graph neural networks for node classification, study the effect of existing post-processing calibration methods, and analyze the influence of model capacity, graph density, and a new loss function on calibration. Further, we propose a topology-aware calibration method that takes the neighboring nodes into account and yields improved calibration compared to baseline methods.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑