比如在两个节点之间观察到事件时(消息传递即通信事件、边增加即交互事件),信息从一个节点流向另一个节点,并相应地影响节点的表示。虽然通信事件(或交互事件)仅在两个节点之间传播局部信息,但关联事件会改变拓扑结构,从而具有更大的全局影响。通过对观察到的局部事件进行编码,从而进一步了解这些事件的局部动态。 3.1 MO...
虽然现有方法已证明在特定环境下非常有效,但它们要么以解耦的方式建模简单的结构和复杂的时间特性,要么使用简单的时间模型(Continuous-Time Dynamic Network Embeddings)。但是,有几个地方表现出结构特性的高度非线性演化与复杂的时间动力学相耦合,有效地建模和学习捕捉此类复杂系统各种动力学特性的信息表示仍然是一个开放的...
现有方法假设图动态演变为一个单一的时间尺度过程。而作者观察到大多数现实世界的图至少表现出两个不同的动态过程,它们在不同的时间尺度上演化——拓扑演化:节点和边的数量预计会随着时间的推移而增长(或缩小),导致图中的结构变化;和节点交互:与节点之间的活动有关,这些节点可能在结构上连接也可能不连接。(ii)如何...
DyRep: Learning Representations over Dynamic Graphs解读 公众号 异度侵入 “ 图结构上的表示学习目前受到极大的关注,由于其在许多任务的优异表现,但目前很多方法针对静态图提出。在这篇文章主要解决两个问题如何编码动态图以及如何高效将动态图编码到低维空间中。作者提出了DyRep框架” 本篇文章由佐治亚理工学院以及Deepm...
How can we effectively encode evolving information over dynamic graphs into low-dimensional representations? In this paper, we propose DyRep, an inductive deep representation learning framework that learns a set of functions to efficiently produce low-dimensional node embeddings that evolves over time. ...
dynamic graphs. (dynamic graph: graph structure, information of edges and nodes vary over time.) hypergraphs (an edge connects two or more nodes). signed graphs (signed edges, which can be positive or negative, e.g. friend/enemy edge). large-scale graphs.Unsupervised...
In: International conference on learning representations. Wang, X., Han, X., Huang, W., Dong, D., & Scott, M.R. (2019). Multi-similarity loss with general pair weighting for deep metric learning. In: Proceedings of the IEEE conference on computer vision and pattern recognition, pp. ...
2.2 Supervised learning over graphs 2.3 Graph convolutional networks 三、Proposed method:GraphSAGE (1)Relation to the Weisfeiler-Lehman Isomorphism Test (1)基于图的无监督损失 ...
Learning graph representations is a fundamental task aimed at capturing various properties of graphs in vector space. The most recent methods learn such representations for static networks. However, real-world networks evolve over time and have varying dynamics. Capturing such evolution is key to predi...
DyRep: Learning Representations over Dynamic Graphs (ICLR, 2019) [paper] Predicting Dynamic Embedding Trajectory in Temporal Interaction Networks (KDD, 2019) [paper][code] Learning Dynamic Context Graphs for Predicting Social Events (KDD, 2019) [paper][code] ...