学习综述论文A Comprehensive Survey on Graph Neural Networks 许多现实世界应用中的图在图结构和图输入这两方面都是动态的。时空图神经网络(STGNNs)在捕捉图的动态性方面占有重要位置。此类别下的方法旨在建模动态节点输入,同时假设连接节点之间的相互依赖性。例如,一个交通网络由路上放置的速度传感器组成,其中边的权重...
Spatial Temporal Modeling Spatial Graph Convolutional Neural Network: Spatial Temporal Modeling 上一节讲的是空间卷积操作,这里重新回到了时间层面,对于t时刻的结点 v_{ti}的邻居结点需要加上在时间点q上的 v_{qj} 满足的条件为 d(v_{ti}, v_{tj}) \leq K ,且满足时间关系 |q-t| \leq [\Gamma/2...
论文:Pre-training Enhanced Spatial-temporal Graph Neural Network for Multivariate Time Series Forecasting 或者是:Pre-training Enhanced Spatial-temporal Graph Neural Network for Multivariate Time Series Forecasting GitHub:https://github.com/zezhishao/STEP KDD 2022的论文。 摘要 多变量时间序列(MTS)预测在广...
论文:Pre-training Enhanced Spatial-temporal Graph Neural Network for Multivariate Time Series Forecasting 或者是:Pre-training Enhanced Spatial-temporal Graph Neural Network for Multivariate Time Series Forecasting GitHub:https://github.com/zezhishao/STEP KDD 2022的论文。 摘要 多变量时间序列(MTS)预测在广...
1)Multi-view Self Attention 在每个视图中,特征在每个时间成对交互。为了便于具体描述,我们以短距离观为例。 然后,我们将时间attention与该值进行加权,得到一个新的特征表示的短距离视图 Global Temporal Attention 我们采用自我注意机制来模拟所有时间步骤的相关性。
In the temporal dimension, we design a novel Gated-Memory Convolutional Neural Network (GMCNN) to capture the non-linear temporal dependencies by controlling the output based on the timing information position. In the spatial dimension, we develop a Multilayer Graph Topological Attention Network (...
DSTAGNN: Dynamic Spatial-Temporal Aware Graph Neural Network for Traffic Flow Forecasting ICML2022 论文地址:https://proceedings.mlr.press/v162/lan22a.html 代码地址:https://github.com/SYLan2019/DSTAGNN 作者:Shiyong Lan, Yitong Ma, Weikang Huang, Wenwu Wang, Hongyu Yang, Pyang Li 一个用于时空...
Zhang et al. [2] 提出了一个View adaptive recurrent neural networks,利用两个LSTM子网络回归骨架的空间旋转参数和空间平移参数,然后将骨架旋转到一个适合行为预测的角度,最后送入主LSTM网络预测行为类别。 Yan et al. [3] 提出了一个 Spatial Temporal Graph Convolutional Networks学习人体骨架序列的时空特征,这是...
Although there are existing works on predicting the future traffic flow, the majority of them have certain limitations on modeling spatial and temporal dependencies. In this paper, we propose a novel spatial temporal graph neural network for traffic flow prediction, which can comprehensively capture ...
HetEmotionNet: Two-Stream Heterogeneous Graph Recurrent Neural Network for Multi-modal Emotion Recognition (MM ’21) 基于生理信号的多媒体刺激下人的情绪研究是一个新兴领域,基于多模态信号的情绪识别已取得重要进展。然而,如何充分利用时空特征之间的互补性进行情感识别,以及如何对多模态信号之间的异质性和相关性...