How Does Bayesian Noisy Self-Supervision Defend Graph Convolutional Networks?
How Does Bayesian Noisy Self-Supervision Defend Graph Convolutional Networks?
复制标题
DOI：
10.1007/s11063-022-10750-8
复制
发表时间：
2022-02
影响因子：
3.1
通讯作者：
Jun Zhuang;M. Hasan
中科院分区：
文献类型：
--
作者：
Jun Zhuang;M. Hasan
近年来，研究表明，与其他当代机器学习模型相比，图卷积网络（GCNs）在节点分类任务上取得了更好的性能。然而，两个潜在的问题威胁到GCNs的鲁棒性，即标签稀缺性和对抗性攻击。从基于自监督的方法、基于对抗的方法和基于检测的方法三个方面加强GCNs的鲁棒性。然而，上述所有方法都很难同时处理这两个问题。本文将噪声监督假设为一种自监督学习方法，并提出了一种新的贝叶斯图噪声自监督模型GraphNS来解决这两个问题。大量的实验表明，GraphNS可以显著增强节点分类，以对抗标签稀缺性和对抗性攻击。这种增强被证明是在四个经典GCNs上推广的，并且优于六个公共图数据集的竞争方法。
In recent years, it has been shown that, compared to other contemporary machine learning models, graph convolutional networks (GCNs) achieve superior performance on the node classification task. However, two potential issues threaten the robustness of GCNs, label scarcity and adversarial attacks. .Intensive studies aim to strengthen the robustness of GCNs from three perspectives, the self-supervision-based method, the adversarial-based method, and the detection-based method. Yet, all of the above-mentioned methods can barely handle both issues simultaneously. In this paper, we hypothesize noisy supervision as a kind of self-supervised learning method and then propose a novel Bayesian graph noisy self-supervision model, namely GraphNS, to address both issues. Extensive experiments demonstrate that GraphNS can significantly enhance node classification against both label scarcity and adversarial attacks. This enhancement proves to be generalized over four classic GCNs and is superior to the competing methods across six public graph datasets.