Graph neural networksSelf-supervised learningEnhance-attentionSession-based recommendation is a challenging task which predicts the next click based on the short-term behavior of anonymous users. Compared to ot
Additionally, multi-task learning can further improve representation learning by training networks simultaneously on related tasks, leading to significant performance improvements. In this paper, we propose three novel self-supervised auxiliary tasks to train graph-based neural network models in a multi-...
网上已有这篇论文的一些讲解,如有讲的不清楚的地方,可移至其他博客。 持续更新中--- Title: Self supervised graph neural network for improving electroencephalographic analysis 作者:Siyi Tang, …
https://github.com/DeepGraphLearning/GraphLoGgithub.com/DeepGraphLearning/GraphLoG 摘要 本文研究了无监督/自监督的全图表示学习,提出了一个局部实例和全局语义学习框架(GraphLoG)。具体来说,除了保留局部相似性,GraphLoG还引入了层次模型来捕捉全局语义集群。 相比于现有算法仅关注不同图实例之间的局部相似性结...
3.节点表示容易收到噪声交互的影响。在本文中作者通过在用户-物品图上引入自监督学习来改善GCN在推荐系统上的准确率和鲁棒性,将其称为Self-supervised Graph Learning(SGL),并应用在LightGCN模型上。SGL是模型无关的,并通过辅助自监督任务来补充监督任务中的信息以达成上述目的。
[arXiv 2023] Structure-Aware Group Discrimination with Adaptive-View Graph Encoder: A Fast Graph Contrastive Learning Framework [paper] [TNNLS 2023] Self-supervised Learning IoT Device Features with Graph Contrastive Neural Network for Device Classification in Social Internet of Things [paper] [TKDE ...
https://github.com/LirongWu/awesome-graph-self-supervised-learning 近些年来,图上的深度学习在各种任务上取得了显著的成功,而这种成功在很大程度上依赖于海量的、精心标注的数据。然而,精确的标注通常非常昂贵和耗时。为了解决这个问题,自监督学习(Self-supervised Learning,SSL)正在成为一种全新的范式,通过精心设计的...
GraphCGAN: Convolutional Graph Neural Network with Generative Adversarial Networks. Transferable Feature Learning on Graphs Across Visual Domains. FASG: Feature Aggregation Self-training GCN for Semi-supervised Node Classification. GL-Disen: Global-Local disentanglement for unsupervised learning of graph-level...
Keywords:Collaborative filtering, Graph Neural Network, Self-supervised Learning, Long-tail Recommendation Abstract 基于用户-项目图的推荐表示学习已经从使用单个ID或交互历史发展到利用高阶邻居。这导致了图卷积网络(GCNs)的成功推荐,如PinSage和LightGCN。尽管有效,但我们认为它们有两个局限性: ...
【Arxiv-2021】【IEEE members/fellows】Graph Self-Supervised Learning: A Survey 核心要点 文章旨在对现有图神经网络的方法进行全面的总结和分类,并给出常用的数据集、评估基准、方法间的性能比较和开源代码链接。图的深度学习的热度与日俱增,但大多数工作集中在(半)监督学习上。对比标签的严重依赖导致模型泛化能力...