图自监督学习和预训练论文推荐

原创

于 2021-10-29 16:39:43 发布 · 519 阅读

4 ·

CC 4.0 BY-SA版权

文章标签：

#机器学习 #图自监督学习 #预训练模型

Self-supervised Learning and Pre-training：研究图的自监督学习和预训练。预训练在NLP任务中大放光采，研究者希望此类技术在图上也能够进行预训练，并期望更好地辅助下游的任务。经典的自监督学习方法如对比学习等。

1.论文名称：Strategies for Pre-training Graph Neural Networks
链接：https://www.aminer.cn/pub/5e5e18eb93d709897ce3ce41
2.论文名称：CommDGI: Community Detection Oriented Deep Graph Infomax
链接：https://www.aminer.cn/pub/5f8ebbb99fced0a24b4e1994
3.论文名称：Inductive Representation Learning on Large Graphs.
链接：https://www.aminer.cn/pub/599c7988601a182cd2648a09
4.论文名称：InfoGraph: Unsupervised and Semi-supervised Graph-Level Representation Learning via Mutual Information Maximization
链接：https://www.aminer.cn/pub/5e5e189a93d709897ce1e760
5.论文名称：GCC: Graph Contrastive Coding for Graph Neural Network Pre-Training
链接：https://www.aminer.cn/pub/5eede0b091e0116a23aafbd3
6.论文名称：Contrastive Multi-View Representation Learning on Graphs
链接：https://www.aminer.cn/pub/5ede0553e06a4c1b26a8419c
7.论文名称：Graph Contrastive Learning with Augmentations