Graph Adversarial Training: Dynamically Regularizing Based on Graph Structure

计算机科学 对抗制 图形 理论计算机科学 稳健性(进化) 人工智能 卷积神经网络 正规化(语言学) 机器学习 生物化学 化学 基因
作者
Fuli Feng,Xiangnan He,Jie Tang,Tat-Seng Chua
出处
期刊:IEEE Transactions on Knowledge and Data Engineering [IEEE Computer Society]
卷期号:33 (6): 2493-2504 被引量:86
标识
DOI:10.1109/tkde.2019.2957786
摘要

Recent efforts show that neural networks are vulnerable to small but intentional perturbations on input features in visual classification tasks. Due to the additional consideration of connections between examples (e.g., articles with citation link tend to be in the same class), graph neural networks could be more sensitive to the perturbations, since the perturbations from connected examples exacerbate the impact on a target example. Adversarial Training (AT), a dynamic regularization technique, can resist the worst-case perturbations on input features and is a promising choice to improve model robustness and generalization. However, existing AT methods focus on standard classification, being less effective when training models on graph since it does not model the impact from connected examples. In this work, we explore adversarial training on graph, aiming to improve the robustness and generalization of models learned on graph. We propose Graph Adversarial Training (GraphAT), which takes the impact from connected examples into account when learning to construct and resist perturbations. We give a general formulation of GraphAT, which can be seen as a dynamic regularization scheme based on the graph structure. To demonstrate the utility of GraphAT, we employ it on a state-of-the-art graph neural network model - Graph Convolutional Network (GCN). We conduct experiments on two citation graphs (Citeseer and Cora) and a knowledge graph (NELL), verifying the effectiveness of GraphAT which outperforms normal training on GCN by 4.51 percent in node classification accuracy. Codes are available via: https://github.com/fulifeng/GraphAT.
最长约 10秒,即可获得该文献文件

科研通智能强力驱动
Strongly Powered by AbleSci AI
更新
PDF的下载单位、IP信息已删除 (2025-6-4)

科研通是完全免费的文献互助平台,具备全网最快的应助速度,最高的求助完成率。 对每一个文献求助,科研通都将尽心尽力,给求助人一个满意的交代。
实时播报
1秒前
吉小聿发布了新的文献求助20
1秒前
2秒前
2秒前
NexusExplorer应助Ruby采纳,获得10
2秒前
zhijiner发布了新的文献求助10
3秒前
Tolerate完成签到 ,获得积分10
4秒前
5秒前
ruby发布了新的文献求助10
5秒前
刘青秀发布了新的文献求助10
5秒前
6秒前
天天快乐应助小小橙采纳,获得10
6秒前
6秒前
冷酷的夜完成签到,获得积分10
7秒前
8秒前
共享精神应助LQX2141采纳,获得10
8秒前
FF完成签到 ,获得积分10
9秒前
优秀爆米花完成签到,获得积分10
10秒前
duanduan发布了新的文献求助10
10秒前
烂番茄完成签到 ,获得积分10
11秒前
追三发布了新的文献求助10
12秒前
GAO发布了新的文献求助10
12秒前
ccq发布了新的文献求助10
12秒前
0376完成签到,获得积分20
12秒前
13秒前
害羞映容发布了新的文献求助10
14秒前
15秒前
15秒前
16秒前
16秒前
Hello应助自由芷雪采纳,获得10
16秒前
悲伤西米露应助苹果笑寒采纳,获得20
16秒前
16秒前
Lucas应助wang佳俊采纳,获得10
17秒前
17秒前
Ava应助禾晏采纳,获得10
18秒前
18秒前
wxy发布了新的文献求助10
18秒前
李健应助ronll采纳,获得10
18秒前
小小橙发布了新的文献求助10
19秒前
高分求助中
A new approach to the extrapolation of accelerated life test data 1000
Picture Books with Same-sex Parented Families: Unintentional Censorship 700
ACSM’s Guidelines for Exercise Testing and Prescription, 12th edition 500
Nucleophilic substitution in azasydnone-modified dinitroanisoles 500
不知道标题是什么 500
Indomethacinのヒトにおける経皮吸収 400
Phylogenetic study of the order Polydesmida (Myriapoda: Diplopoda) 370
热门求助领域 (近24小时)
化学 材料科学 医学 生物 工程类 有机化学 生物化学 物理 内科学 纳米技术 计算机科学 化学工程 复合材料 遗传学 基因 物理化学 催化作用 冶金 细胞生物学 免疫学
热门帖子
关注 科研通微信公众号,转发送积分 3976177
求助须知:如何正确求助?哪些是违规求助? 3520366
关于积分的说明 11202745
捐赠科研通 3256847
什么是DOI,文献DOI怎么找? 1798509
邀请新用户注册赠送积分活动 877704
科研通“疑难数据库(出版商)”最低求助积分说明 806516