arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2205.02468cs.LGcs.AI
收藏

Alignahead:基于图神经网络的在线跨层知识提取

Alignahead: Online Cross-Layer Knowledge Extraction on Graph Neural Networks

Zhejiang University(浙江大学);ZJU-Bangsun Joint Research Center(浙大邦生联合研究中心);Shanghai Institute for Advanced Study of Zhejiang University(浙江大学上海高等研究院)

Jiongyu Guo, Defang Chen, Can Wang

摘要:

现有的图神经网络(GNN)上的知识蒸馏方法几乎是离线的,其中学生模型从强大的教师模型中提取知识以提高其性能。然而,一个预先训练的教师模型并不总是访问由于培训成本,隐私等,在本文中,我们提出了一种新的在线知识蒸馏框架来解决这个问题。具体来说,每个学生GNN模型在交替训练过程中从另一个同时训练的对应物中学习提取的局部结构。我们进一步开发了一种跨层蒸馏策略,通过将一个学生层与另一个学生模型的不同深度的层对齐,理论上使结构信息分散在所有层上。在PPI、Coauthor-CS/Physics和Amazon-Computer/Photo等五个数据集上的实验结果表明,在没有预先训练的教师模型的监督下,学生的成绩在我们的协作训练框架中得到了持续的提高。此外,我们还发现,我们的alignahead技术可以加快模型的收敛速度,其有效性可以通过增加训练中的学生人数普遍提高。代码可用:https://github.com/GuoJY-eatsTG/Alignahead

英文摘要:

Existing knowledge distillation methods on graph neural networks (GNNs) are almost offline, where the student model extracts knowledge from a powerful teacher model to improve its performance. However, a pre-trained teacher model is not always accessible due to training cost, privacy, etc. In this paper, we propose a novel online knowledge distillation framework to resolve this problem. Specifically, each student GNN model learns the extracted local structure from another simultaneously trained counterpart in an alternating training procedure. We further develop a cross-layer distillation strategy by aligning ahead one student layer with the layer in different depth of another student model, which theoretically makes the structure information spread over all layers. Experimental results on five datasets including PPI, Coauthor-CS/Physics and Amazon-Computer/Photo demonstrate that the student performance is consistently boosted in our collaborative training framework without the supervision of a pre-trained teacher model. In addition, we also find that our alignahead technique can accelerate the model convergence speed and its effectiveness can be generally improved by increasing the student numbers in training. Code is available: https://github.com/GuoJY-eatsTG/Alignahead

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。

↑